Research ArticleA Longitudinal Low Dose 120583CT Analysis of Bone Healing in MiceA Pilot Study

Lu-Zhao Di1 Vanessa Couture1 Eacutelisabeth Leblanc12 Yasaman Alinejad1

Jean-Franccedilois Beaudoin13 Roger Lecomte134 Franccedilois Berthod56 Nathalie Faucheux17

Freacutedeacuteric Balg12 and Guillaume Grenier12

1 Research Center of CHUS (CRCHUS) 3001 12th Avenue North Sherbrooke QC Canada J1H 5N42Department of Orthopedic Surgery Faculty of Medicine and Health Sciences University of Sherbrooke 3001 12th Avenue NorthSherbrooke QC Canada J1H 5N4

3 Sherbrooke Molecular Imaging Centre (CIMS) CRCHUS 3001 12th Avenue North Sherbrooke QC Canada J1H 5N44Department of Nuclear Medicine and Radiobiology Faculty of Medicine and Health Sciences University of Sherbrooke 300112th Avenue North Sherbrooke QC Canada J1H 5N4

5 Laboratoire drsquoOrganogenese Experimentale (LOEX) Research Center of CHUQ Enfant-Jesus Hospital 1401 18th Street Quebec CityQC Canada G1J 1Z4

6Department of Surgery Faculty of Medicine Laval University 2325 Universite Street Quebec City QC Canada G1V 0A67Department of Chemical and Biotechnological Engineering Faculty of Engineering University of Sherbrooke2500 Universite Boulevard Sherbrooke QC Canada J1K 2R1

Correspondence should be addressed to Guillaume Grenier guillaumegrenierusherbrookeca

Received 11 June 2014 Accepted 13 October 2014 Published 6 November 2014

Academic Editor Panagiotis Korovessis

Copyright copy 2014 Lu-Zhao Di et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Low dose microcomputed tomography (120583CT) is a recently matured technique that enables the study of longitudinal bone healingand the testing of experimental treatments for bone repairThis imaging technique has been used for studying craniofacial repair inmice but not in an orthopedic contextThis is mainly due to the size of the defects (approximately 10mm) in long bone which healrapidly andmay thus negatively impact the assessment of the effectiveness of experimental treatmentsWe developed a longitudinallow dose 120583CT scan analysis method combined with a new image segmentation and extraction software using Hounsfield unit (HU)scores to quantitatively monitor bone healing in small femoral cortical defects in live mice We were able to reproducibly quantifybone healing longitudinally over time with three observers We used high speed intramedullary reaming to prolong healing inorder to circumvent the rapid healing typical of small defects Bone healing prolongation combined with 120583CT imaging to studysmall bone defects in live mice thus shows potential as a promising tool for future preclinical research on bone healing

1 Introduction

Bone healing is a constantly growing field of research withimportant clinical implications Precise and effective moni-toring of bone healing is thus essential Histological prepa-rations are traditionally used to assess tissue compositionand bone repair processes However they are limited to two-dimensional analyses of three-dimensional (3D) structuresand semiquantitative scoring systems Biomechanical testingis also a reliable quantitative method for assessing bonerepair but is limited to endpoint analyses as are histological

preparations [1] Because of high inter- and intrasubjectvariability such studies involving several time points requirelarge numbers of animals which increases the cost of labor-intensive experiments [2]

The use of 120583CT in experimental medicine has increasedexponentially over the past few years because this imagingtechnique provides reliable and highly reproducible results[3]120583CT imaging enables both noninvasive tissue-preservingimaging and quantitative 3D morphometry of bone struc-tures notably of woven bone [4 5] and has thus rapidlyevolved to become the gold standard for microarchitecture

Hindawi Publishing CorporationAdvances in OrthopedicsVolume 2014 Article ID 791539 9 pageshttpdxdoiorg1011552014791539

2 Advances in Orthopedics

(a) (b)

(c) (d)

Figure 1 Surgical procedure for creating diaphyseal femoral cortical defects and intramedullary reaming The femur was exposed (a) andthe quadriceps was reclined by everting the patella (b) A cortical defect was created in the distal midshaft portion of the femur using a drillbit (c) Reaming was performed by drilling through the intercondylar notch of the femur (d)

analyses [6 7] Longitudinal low dose 120583CT analyses havebeen used to study treatments for enhancing the healing ofcalvarial defects in murine models [8ndash11] and to quantifytumor-induced osteolysis [12] However 120583CT data collectionoften requires ex vivo specimens meaning that the animalshave to be sacrificed [6] While a few investigators havereported using low dose 120583CT to study bone architecturein vivo [13ndash15] none has used longitudinal 120583CT scans toevaluate the healing of drill hole cortical defects in micemainly because spontaneous healing occurs quickly makingit difficult to assess the effect of treatments on bone healing[16] To slow down spontaneous bone healing Gao et alirradiated the limbs of mice to reduce the contributionof endogenous cells to bone healing [17] while effectiveirradiation impairs nonstem cell therapies such as growthfactors that stimulate the recruitment and differentiation ofthe endogenous progenitors required for bone repair

In the present study we used low dose 120583CT imaging toperform longitudinal quantitative analyses of bone healingof femoral cortical defects in mice To assess bone healingwe performed longitudinal 120583CT scans of live animals usinga dedicated small-animal low dose 120583CT scanner and ana-lyzed the images using custom semiautomated segmenta-tion and coregistration software Despite creating noncriticalsize femoral defects our results enabled us to show thatintramedullary reaming is effective in slowing down spon-taneous bone healing and significantly increasing the timerequired for the defect to heal In summary the combination

of low dose 120583CT scan imaging of live animals and bonereaming appears to be a valuable tool for evaluating thekinetics of long bone healing following treatments withdrugs biomaterials andor cells in mice

2 Materials and Methods

21 Animals Skeletally mature 12- to 16-week-old male CD1(Charles River Saint-Constant QC Canada)mice were usedThis study was carried out in strict accordance with therecommendations in the Guide for the Care and Use ofLaboratory Animals of the National Institutes of HealthThe protocol was approved by the Committee on the Ethicsof Animal Experiments of the University of Sherbrooke(Permit Number 141-11B) Prior to surgery the mice receiveda narcotic (Buprenorphine 005ndash01mgkg) to control pain aswell as a preventive dose of antibiotic (penicillinG 60000U)

22 Surgical Procedure for Creating Femoral Cortical Defectsand Reaming the Medullar Cavity A diaphyseal corticaldefect of the anterodistal femur was created using a modifieddrill hole technique [18 19] with or without intramedullaryreaming as shown in Figure 1 Briefly the mice were anes-thetized (isoflurane) their limbs were shaved and the skinwas disinfected using a 05 chlorhexidine solution (Dexidin05 Atlas LaboratoryMontreal QC Canada) A skin incisionwas made to expose the medial quadriceps and knee Amedial parapatellar approach was then made to the knee and

Advances in Orthopedics 3

was extended proximally through the quadriceps to exposethe distal femur by lateral patellar eversion A 36410158401015840 drill bit(Model 865 Dremel Mount Prospect IL USA) was used tocreate a 11mm-diameter circular defect in the anterior cortexof both femurs 4mm above the femoral condylesThe defectswere irrigated with normal saline to remove bone residuesIntramedullary reaming was performed on one femur bydrilling a 13210158401015840-hole through the intercondylar notch andthen reaming the full length of the medullar cavity Thereamer was kept in the medullar cavity at high speed for 10 s

The defects were then covered by reclining the quadricepsback to its original position and repositioning the patellaThe incision was closed using nonabsorbable sutures Thesurgerywas performedby an orthopedic surgeonusing sterilematerials in aseptic conditions The mice were used forlive 120583CT imaging andor were euthanized for histologicalexaminations

23 Low Dose Microcomputed Tomography (120583CT) Scan Imag-ing Low dose 120583CT scan imaging was performed using aGamma Medica Triumph X-O small-animal CT scannercomposed of a 40W X-ray tube with a 75 120583m focal spotdiameter and a 2240 times 2368CsI flat panel X-ray detectorThe detector pixel size was 50 120583m and a 2 times 2 pixel binningscheme was used Scans were performed at 60 kVp and230 120583A using 512 projections in fly mode to reduce exposurewith a 5920mm FOV Images were reconstructed using thegeneral purpose reconstruction FBP kernel from the GammaMedica software over an isotropic 116120583m voxel spacinggrid providing a voxel volume of 156 times 10minus3mm3 Air andwater phantoms (15mL Falcon tubes) were scanned for eachimaging session to allow normalization in HU The live micewere scanned once prior to surgery as a control and thenat days 0 (immediately after surgery) 7 14 21 30 35 4457 and 78 after surgery To minimize movement the micewere isoflurane-sedated during the 120583CT imaging The doseof radiation administered was 117 cGy per scan as measuredby dosimeter

24 Software to Quantify Bone Healing Custom softwarewas written to automatically calculate the size of the cor-tical defects The 120583CT scans consisted of 512 individualimage slices stacked in a RAW file and converted to theNIFTI format using Fiji freeware (ImageJ Version 142qNational Institutes of Health USA) [20] ITK-SNAP freeware(httpwwwitksnaporg) was used to segment structures [21]and manually delineate 3D ROI on the cortical defects inthe postsurgery (day 0) scans Tracing an irregular anatomiccontour adjacent to the cortical surface was deemed the bestapproach to create the 3D ROI We used software providedby the Image Analysis and Visualization Platform (PAVISherbrookeQC) to virtually extract align and save images ofeachmouse femur to smaller volumesThe virtually extractedfemurs were then registered to the day 0 postsurgery refer-ence femur using FSL Flirt [22 23] Since we did not want thespatial information of the bones to be lost we parameterizedFlirt to use a rigid body transformation with 6 degreesof freedom After registration a linear transformation was

applied to the voxel signal intensities to convert them to theHU [24] scale using water for calibration This allowed thesignals from different scans made at different time pointsto be calibrated The HU scale makes it possible to conductanalyses of absolute voxel signal intensities in order todetermine the material density in each voxel (bone musclewater etc) We then calculated the kinetics of cortical defecthealing using the previously determined 3D ROI whichbecame relevant for each time point after registration Thevolumes and sizes of the cortical defects in treated and controlfemurs for each mouse were then quantified over time aswere their densities based on the HU scale The HU rangesfor empty or soft tissue reaction (eg hematoma) (minus1000 to+800HU) woven bone (+1200 to +1900HU) and compactbone (gt2700HU comparable to intact bone) were basedon published results from other groups [25ndash27] as well ason results from specific anatomical sites with known tissuecompositions

25 Statistics To calculate the sample size required we usedthe following repeated sample size estimation formula (119899 =2 + 119862(119904119889)

2) where 119904 is the standard deviation and 119889 is the

difference measured with a 119862 of 785 a power of 80 andan alpha of 005 The difference 119889 was measured between thevolumes of the cortical deficits in the left and right femurson postoperative day 29 when the dissimilarity between thetwo femurs was themostmarkedThe standard deviation wasalso calculated using the results on postoperative day 29 Thestandard deviation was 0255 and the measured differencewas 036The sample size required for a significant result wasthus at least 6 mice

One outcome of the study was the assessment of theagreement among three observers with respect to their3D ROI determinations for the same 120583CT scan This wasrepresented by a Bland-Altman plot [28] Graphically themean differences and the 95 confidence intervals (CI)were represented An intraclass coefficient (ICC) was alsoused to assess their reliability and validity The ICC for theinterobserver agreement was obtained by comparing blindmeasurements of the three observers For clinical studiesa Cronbach alpha value over 085 is considered very goodwhile a value over 09 is considered ldquoidealrdquo [29]

The main outcome of the present study was to assessthe differences in 3D ROI voxel HU cortical defect volumesbetween the reamed and unreamed femurs relative to therespective postoperative day 0 volumes The comparisonof means between reamed and unreamed samples at thesame time point was analyzed using two-way ANOVA andSidakrsquosmultiple comparisons test119875 values less than 005wereconsidered significant

Statistical analyses were performed using SPSS v2000and the results were graphed using Prism v50

3 Results

31 3D ROI Analysis of Interobserver Agreement Since initial3D ROI segmentation may be subjective and thus overesti-mate or underestimate healing we verified the interobserver

4 Advances in Orthopedics

Average (voxels)

500 1000 1500 20000

200

400

IInte

robs

erve

r diff

eren

ces (

voxe

ls)

Defect alone

A-BB-CC-ABias 007SD bias 1132

minus400

minus200

(a)

500 1000 1500 20000

200

400

Average (voxels)

Defect and reamed

Inte

robs

erve

r diff

eren

ces (

voxe

ls)

Bias minus007SD bias 1233

minus400

minus200

A-BB-CC-A

(b)

Figure 2 Interobserver agreement in the determination of 3DROIABland-Altmanplotwas used to graph differences in the absolute numberof voxels per 3D ROI as a function of the average size of the defect This made it possible to assess the agreement between three observers inthe determination of the 3D ROI Eighteen 120583CT images from (a) unreamed femurs with a defect and (b) reamed femurs with a defect wereblindly analyzed The two-by-two comparison of the three observers (A-B B-C and C-A) showed a 95 confidence interval from minus2265 to+2264 and from minus2459 to +2471 for unreamed femurs with a defect and reamed femurs with a defect respectively An interclass correlation(ICC) was used to compare interobserver variability The ICC scores were 0864 (very good 119875 lt 00001) and 0905 (ideal 119875 lt 00001) forthe unreamed femurs with a defect and the reamed femurs with a defect respectively

agreement which was represented using Bland-Altman plots(Figure 2) We used three observers (A-B-C) to blindlydetermine the 3DROI of the defectsThe results are presentedas the difference in the absolute number of voxels per3D ROI as a function of the average size of the defectOf the 36 scans 18 were of femurs with a defect but noreaming (Figure 2(a)) and 18 were of femurs with a defectand reaming (Figure 2(b)) The two-by-two comparison ofthe three observers (A-B B-C and C-A) showed that the 95confidence interval of the evaluations of the observers rangedfromminus2265 to +2264 (219plusmn 50) and fromminus2459 to +2471(201 plusmn 32) for defects without reaming and defects withreaming respectively

To determine whether the interobserver agreement wassignificant we calculated an interclass correlation (ICC)which allowed us to compare interobserver variability for agiven sampleThe ICC was 0864 (very good 119875 lt 00001) forthe defect without reaming and 0905 (ideal 119875 lt 00001) forthe defect with reaming indicating very good interobserveragreement In addition there was no significant differencebetween the 3D ROI of the defect without reaming and thedefect with reaming

32 Longitudinal Low Dose 120583CT Analysis of Bone HealingWe determined the difference in the kinetics of bone healingbetween reamed and unreamed femurs by calculating thenumber of voxels in nonorganized substrate woven boneand compact bone Figure 3(a) shows reconstructed images

of anteroposterior views of the femurs with their defects ondays 0 and 21 as well as on day 57 after surgery when themostsignificant variations in tissue composition of the reamed andunreamed femurs were observed In addition to visualizingthe registration of the femurs at different times it is possibleto identify the position of the 3D ROIs overlaid on the defectsalong different views (Figure 3(b)) Cross-sectional views ofthe femurs can also be seen in which the 3D ROI is apparent(Figure 3(c)) In both representations the voxels of the 3DROI are color-coded to differentiate between the variouscomponents in the defect that is emptysoft tissue reaction(red) woven bone (blue) and compact bone comparable tointact bone (green) Figures 3(b) and 3(c) show that wovenbone and compact bone formation is delayed in the reamedpreparation compared to the unreamed preparation

We used our custom software to verify the filling kineticsof the defects (Figure 4(a)) Graphing the percentage of voxelsin the minus1000 to +800HU range within the 3D ROI over timerevealed that reaming caused a significant delay in healingIt took 11 and 16 days after surgery for 50 healing to occurin the unreamed and reamed femurs respectively while 75healing took approximately 16 and 36 days respectively Byday 57 there was no significant difference in healing betweenthe two surgical procedures

We also assessed the formation of woven bone by graph-ing the percentage of voxels in the +1200 to +1900HU rangewithin the 3D ROI as a function of time (Figure 4(b)) Thepercentage of voxels related to woven bone was significantly

Advances in Orthopedics 5

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

2mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(a)

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

Axi

alC

oron

alSa

gitta

lA

xial

Cor

onal

Sagi

ttal

Axi

alC

oron

alSa

gitta

l

1mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(b)

Defectand reamedDefect alone

Day

0D

ay 2

1D

ay 5

7

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

1mm

(c)

Figure 3 Bone healing monitoring as a function of time using low dose 120583CT scans of live mice (a) Representative views of 3D 120583CTreconstructed images of femurs on days 0 21 and 57 after surgery Variations in the density of the 3D ROI (determined at day 0 after surgery)make it possible to estimate the tissue composition (b) 3D ROI can also be observed in axial coronal and sagittal views (c) Representativecross-sectional views of the femur in which the 3D ROI is apparent in one image (slice) The colored voxels make it possible to see changesin defect composition as a function of time where most of the voxels were red on day 0 and blue on day 28 with an increasing proportionof green on day 57 Nonorganized substrate (red minus1000 to +800HU) woven bone (blue +1200 to +1900HU) and compact bone (greengt2700HU)

higher in the unreamed preparations with the percent-age reaching a plateau around day 29 of 40 and 29for the unreamed and reamed preparations respectivelyInterestingly the percentages decreased to 20 by day 57when no significant difference between the preparations wasobserved

We investigated the formation of compact bone by graph-ing the percentage of voxels in the 3D ROI with gt2700HU

as a function of time (Figure 4(c)) The percentage of voxelsrelated to dense bone increased significantly faster in theunreamed femurs than in the reamed femurs This differencewas observed by day 35 when the percentage reached 30for the unreamed femurs but only 8 for the reamed femursThe percentage increased until day 57 reaching nearly 40for the unreamed femurs and nearly 22 for the reamedfemurs

6 Advances in Orthopedics

0 20 40 60 800

20

40

60

80

100

Voxe

l (

)

Defect aloneDefect and reamed

Time (days)

lowastlowastlowast

lowast

lowast

lowast lowast

Emptysoft tissue reaction (minus1000 to +800HU)

(a)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowast

Woven bone (+1200 to +1900 HU)

(b)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowastlowastlowastlowastlowast

lowastlowastlowast

lowastlowastlowast

Compact bone (gt+2700 HU)

(c)

Figure 4 Bone healing as a function of time for a defect without reaming and a defect with reaming (a) The results were quantified andplotted as a percentage of voxels representing nonorganized material as a function of time for unreamed femurs with a defect and reamedfemurs with a defect The percentages of (b) woven and (c) dense bone formation as a function of time were also graphed The results wereobtained from nine mice (119899 = 9 lowast119875 lt 005 lowastlowastlowast119875 lt 00001)

4 Discussion

120583CT is a precise tool for bone imaging and in recent decadesit has been shown to be the gold standard for studying bonestructure [30] It has been shown to be a relevant tool forcraniofacial studies in small rodents in which it is possibleto create critical size bone or nonhealing defects [8ndash11]Unfortunately such critical size defects cannot be created inthe long bones of mice without using nails or other materialsthat can impair 120583CT imaging [31 32] On the other handsmall noncritical size bone defects heal spontaneously andrapidly which may hinder the effectiveness of experimentaltreatments [16] To widen this ldquotherapeuticalrdquo window wereamed the endosteum of the medullar region of the bone

The aim of our study was to evaluate the utility of usinglongitudinal low dose 120583CT scans to assess the healing ofnoncritical size defects (1mm) in long bones in mice and toinvestigate the effect of intramedullary reaming on healingtimes

Longitudinal studies of small cortical defects using lowdose 120583CT scan are challenging for many reasons First bonedefects created using small drill bits are subject to operatorbias due to differences in the angle at which the drill bitenters the femur and due to micromovements once the bit isfully introduced This bias must be taken into account giventhe size of the defects (1mm) which are much smaller thancalvarial defects which can be up to 5mm in diameter inmice [8] Second small defects are assessed using multiple

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

2 Advances in Orthopedics

(a) (b)

(c) (d)

Figure 1 Surgical procedure for creating diaphyseal femoral cortical defects and intramedullary reaming The femur was exposed (a) andthe quadriceps was reclined by everting the patella (b) A cortical defect was created in the distal midshaft portion of the femur using a drillbit (c) Reaming was performed by drilling through the intercondylar notch of the femur (d)

analyses [6 7] Longitudinal low dose 120583CT analyses havebeen used to study treatments for enhancing the healing ofcalvarial defects in murine models [8ndash11] and to quantifytumor-induced osteolysis [12] However 120583CT data collectionoften requires ex vivo specimens meaning that the animalshave to be sacrificed [6] While a few investigators havereported using low dose 120583CT to study bone architecturein vivo [13ndash15] none has used longitudinal 120583CT scans toevaluate the healing of drill hole cortical defects in micemainly because spontaneous healing occurs quickly makingit difficult to assess the effect of treatments on bone healing[16] To slow down spontaneous bone healing Gao et alirradiated the limbs of mice to reduce the contributionof endogenous cells to bone healing [17] while effectiveirradiation impairs nonstem cell therapies such as growthfactors that stimulate the recruitment and differentiation ofthe endogenous progenitors required for bone repair

In the present study we used low dose 120583CT imaging toperform longitudinal quantitative analyses of bone healingof femoral cortical defects in mice To assess bone healingwe performed longitudinal 120583CT scans of live animals usinga dedicated small-animal low dose 120583CT scanner and ana-lyzed the images using custom semiautomated segmenta-tion and coregistration software Despite creating noncriticalsize femoral defects our results enabled us to show thatintramedullary reaming is effective in slowing down spon-taneous bone healing and significantly increasing the timerequired for the defect to heal In summary the combination

of low dose 120583CT scan imaging of live animals and bonereaming appears to be a valuable tool for evaluating thekinetics of long bone healing following treatments withdrugs biomaterials andor cells in mice

2 Materials and Methods

21 Animals Skeletally mature 12- to 16-week-old male CD1(Charles River Saint-Constant QC Canada)mice were usedThis study was carried out in strict accordance with therecommendations in the Guide for the Care and Use ofLaboratory Animals of the National Institutes of HealthThe protocol was approved by the Committee on the Ethicsof Animal Experiments of the University of Sherbrooke(Permit Number 141-11B) Prior to surgery the mice receiveda narcotic (Buprenorphine 005ndash01mgkg) to control pain aswell as a preventive dose of antibiotic (penicillinG 60000U)

22 Surgical Procedure for Creating Femoral Cortical Defectsand Reaming the Medullar Cavity A diaphyseal corticaldefect of the anterodistal femur was created using a modifieddrill hole technique [18 19] with or without intramedullaryreaming as shown in Figure 1 Briefly the mice were anes-thetized (isoflurane) their limbs were shaved and the skinwas disinfected using a 05 chlorhexidine solution (Dexidin05 Atlas LaboratoryMontreal QC Canada) A skin incisionwas made to expose the medial quadriceps and knee Amedial parapatellar approach was then made to the knee and

Advances in Orthopedics 3

was extended proximally through the quadriceps to exposethe distal femur by lateral patellar eversion A 36410158401015840 drill bit(Model 865 Dremel Mount Prospect IL USA) was used tocreate a 11mm-diameter circular defect in the anterior cortexof both femurs 4mm above the femoral condylesThe defectswere irrigated with normal saline to remove bone residuesIntramedullary reaming was performed on one femur bydrilling a 13210158401015840-hole through the intercondylar notch andthen reaming the full length of the medullar cavity Thereamer was kept in the medullar cavity at high speed for 10 s

The defects were then covered by reclining the quadricepsback to its original position and repositioning the patellaThe incision was closed using nonabsorbable sutures Thesurgerywas performedby an orthopedic surgeonusing sterilematerials in aseptic conditions The mice were used forlive 120583CT imaging andor were euthanized for histologicalexaminations

23 Low Dose Microcomputed Tomography (120583CT) Scan Imag-ing Low dose 120583CT scan imaging was performed using aGamma Medica Triumph X-O small-animal CT scannercomposed of a 40W X-ray tube with a 75 120583m focal spotdiameter and a 2240 times 2368CsI flat panel X-ray detectorThe detector pixel size was 50 120583m and a 2 times 2 pixel binningscheme was used Scans were performed at 60 kVp and230 120583A using 512 projections in fly mode to reduce exposurewith a 5920mm FOV Images were reconstructed using thegeneral purpose reconstruction FBP kernel from the GammaMedica software over an isotropic 116120583m voxel spacinggrid providing a voxel volume of 156 times 10minus3mm3 Air andwater phantoms (15mL Falcon tubes) were scanned for eachimaging session to allow normalization in HU The live micewere scanned once prior to surgery as a control and thenat days 0 (immediately after surgery) 7 14 21 30 35 4457 and 78 after surgery To minimize movement the micewere isoflurane-sedated during the 120583CT imaging The doseof radiation administered was 117 cGy per scan as measuredby dosimeter

24 Software to Quantify Bone Healing Custom softwarewas written to automatically calculate the size of the cor-tical defects The 120583CT scans consisted of 512 individualimage slices stacked in a RAW file and converted to theNIFTI format using Fiji freeware (ImageJ Version 142qNational Institutes of Health USA) [20] ITK-SNAP freeware(httpwwwitksnaporg) was used to segment structures [21]and manually delineate 3D ROI on the cortical defects inthe postsurgery (day 0) scans Tracing an irregular anatomiccontour adjacent to the cortical surface was deemed the bestapproach to create the 3D ROI We used software providedby the Image Analysis and Visualization Platform (PAVISherbrookeQC) to virtually extract align and save images ofeachmouse femur to smaller volumesThe virtually extractedfemurs were then registered to the day 0 postsurgery refer-ence femur using FSL Flirt [22 23] Since we did not want thespatial information of the bones to be lost we parameterizedFlirt to use a rigid body transformation with 6 degreesof freedom After registration a linear transformation was

applied to the voxel signal intensities to convert them to theHU [24] scale using water for calibration This allowed thesignals from different scans made at different time pointsto be calibrated The HU scale makes it possible to conductanalyses of absolute voxel signal intensities in order todetermine the material density in each voxel (bone musclewater etc) We then calculated the kinetics of cortical defecthealing using the previously determined 3D ROI whichbecame relevant for each time point after registration Thevolumes and sizes of the cortical defects in treated and controlfemurs for each mouse were then quantified over time aswere their densities based on the HU scale The HU rangesfor empty or soft tissue reaction (eg hematoma) (minus1000 to+800HU) woven bone (+1200 to +1900HU) and compactbone (gt2700HU comparable to intact bone) were basedon published results from other groups [25ndash27] as well ason results from specific anatomical sites with known tissuecompositions

25 Statistics To calculate the sample size required we usedthe following repeated sample size estimation formula (119899 =2 + 119862(119904119889)

2) where 119904 is the standard deviation and 119889 is the

difference measured with a 119862 of 785 a power of 80 andan alpha of 005 The difference 119889 was measured between thevolumes of the cortical deficits in the left and right femurson postoperative day 29 when the dissimilarity between thetwo femurs was themostmarkedThe standard deviation wasalso calculated using the results on postoperative day 29 Thestandard deviation was 0255 and the measured differencewas 036The sample size required for a significant result wasthus at least 6 mice

One outcome of the study was the assessment of theagreement among three observers with respect to their3D ROI determinations for the same 120583CT scan This wasrepresented by a Bland-Altman plot [28] Graphically themean differences and the 95 confidence intervals (CI)were represented An intraclass coefficient (ICC) was alsoused to assess their reliability and validity The ICC for theinterobserver agreement was obtained by comparing blindmeasurements of the three observers For clinical studiesa Cronbach alpha value over 085 is considered very goodwhile a value over 09 is considered ldquoidealrdquo [29]

The main outcome of the present study was to assessthe differences in 3D ROI voxel HU cortical defect volumesbetween the reamed and unreamed femurs relative to therespective postoperative day 0 volumes The comparisonof means between reamed and unreamed samples at thesame time point was analyzed using two-way ANOVA andSidakrsquosmultiple comparisons test119875 values less than 005wereconsidered significant

Statistical analyses were performed using SPSS v2000and the results were graphed using Prism v50

3 Results

31 3D ROI Analysis of Interobserver Agreement Since initial3D ROI segmentation may be subjective and thus overesti-mate or underestimate healing we verified the interobserver

4 Advances in Orthopedics

Average (voxels)

500 1000 1500 20000

200

400

IInte

robs

erve

r diff

eren

ces (

voxe

ls)

Defect alone

A-BB-CC-ABias 007SD bias 1132

minus400

minus200

(a)

500 1000 1500 20000

200

400

Average (voxels)

Defect and reamed

Inte

robs

erve

r diff

eren

ces (

voxe

ls)

Bias minus007SD bias 1233

minus400

minus200

A-BB-CC-A

(b)

Figure 2 Interobserver agreement in the determination of 3DROIABland-Altmanplotwas used to graph differences in the absolute numberof voxels per 3D ROI as a function of the average size of the defect This made it possible to assess the agreement between three observers inthe determination of the 3D ROI Eighteen 120583CT images from (a) unreamed femurs with a defect and (b) reamed femurs with a defect wereblindly analyzed The two-by-two comparison of the three observers (A-B B-C and C-A) showed a 95 confidence interval from minus2265 to+2264 and from minus2459 to +2471 for unreamed femurs with a defect and reamed femurs with a defect respectively An interclass correlation(ICC) was used to compare interobserver variability The ICC scores were 0864 (very good 119875 lt 00001) and 0905 (ideal 119875 lt 00001) forthe unreamed femurs with a defect and the reamed femurs with a defect respectively

agreement which was represented using Bland-Altman plots(Figure 2) We used three observers (A-B-C) to blindlydetermine the 3DROI of the defectsThe results are presentedas the difference in the absolute number of voxels per3D ROI as a function of the average size of the defectOf the 36 scans 18 were of femurs with a defect but noreaming (Figure 2(a)) and 18 were of femurs with a defectand reaming (Figure 2(b)) The two-by-two comparison ofthe three observers (A-B B-C and C-A) showed that the 95confidence interval of the evaluations of the observers rangedfromminus2265 to +2264 (219plusmn 50) and fromminus2459 to +2471(201 plusmn 32) for defects without reaming and defects withreaming respectively

To determine whether the interobserver agreement wassignificant we calculated an interclass correlation (ICC)which allowed us to compare interobserver variability for agiven sampleThe ICC was 0864 (very good 119875 lt 00001) forthe defect without reaming and 0905 (ideal 119875 lt 00001) forthe defect with reaming indicating very good interobserveragreement In addition there was no significant differencebetween the 3D ROI of the defect without reaming and thedefect with reaming

32 Longitudinal Low Dose 120583CT Analysis of Bone HealingWe determined the difference in the kinetics of bone healingbetween reamed and unreamed femurs by calculating thenumber of voxels in nonorganized substrate woven boneand compact bone Figure 3(a) shows reconstructed images

of anteroposterior views of the femurs with their defects ondays 0 and 21 as well as on day 57 after surgery when themostsignificant variations in tissue composition of the reamed andunreamed femurs were observed In addition to visualizingthe registration of the femurs at different times it is possibleto identify the position of the 3D ROIs overlaid on the defectsalong different views (Figure 3(b)) Cross-sectional views ofthe femurs can also be seen in which the 3D ROI is apparent(Figure 3(c)) In both representations the voxels of the 3DROI are color-coded to differentiate between the variouscomponents in the defect that is emptysoft tissue reaction(red) woven bone (blue) and compact bone comparable tointact bone (green) Figures 3(b) and 3(c) show that wovenbone and compact bone formation is delayed in the reamedpreparation compared to the unreamed preparation

We used our custom software to verify the filling kineticsof the defects (Figure 4(a)) Graphing the percentage of voxelsin the minus1000 to +800HU range within the 3D ROI over timerevealed that reaming caused a significant delay in healingIt took 11 and 16 days after surgery for 50 healing to occurin the unreamed and reamed femurs respectively while 75healing took approximately 16 and 36 days respectively Byday 57 there was no significant difference in healing betweenthe two surgical procedures

We also assessed the formation of woven bone by graph-ing the percentage of voxels in the +1200 to +1900HU rangewithin the 3D ROI as a function of time (Figure 4(b)) Thepercentage of voxels related to woven bone was significantly

Advances in Orthopedics 5

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

2mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(a)

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

Axi

alC

oron

alSa

gitta

lA

xial

Cor

onal

Sagi

ttal

Axi

alC

oron

alSa

gitta

l

1mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(b)

Defectand reamedDefect alone

Day

0D

ay 2

1D

ay 5

7

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

1mm

(c)

Figure 3 Bone healing monitoring as a function of time using low dose 120583CT scans of live mice (a) Representative views of 3D 120583CTreconstructed images of femurs on days 0 21 and 57 after surgery Variations in the density of the 3D ROI (determined at day 0 after surgery)make it possible to estimate the tissue composition (b) 3D ROI can also be observed in axial coronal and sagittal views (c) Representativecross-sectional views of the femur in which the 3D ROI is apparent in one image (slice) The colored voxels make it possible to see changesin defect composition as a function of time where most of the voxels were red on day 0 and blue on day 28 with an increasing proportionof green on day 57 Nonorganized substrate (red minus1000 to +800HU) woven bone (blue +1200 to +1900HU) and compact bone (greengt2700HU)

higher in the unreamed preparations with the percent-age reaching a plateau around day 29 of 40 and 29for the unreamed and reamed preparations respectivelyInterestingly the percentages decreased to 20 by day 57when no significant difference between the preparations wasobserved

We investigated the formation of compact bone by graph-ing the percentage of voxels in the 3D ROI with gt2700HU

as a function of time (Figure 4(c)) The percentage of voxelsrelated to dense bone increased significantly faster in theunreamed femurs than in the reamed femurs This differencewas observed by day 35 when the percentage reached 30for the unreamed femurs but only 8 for the reamed femursThe percentage increased until day 57 reaching nearly 40for the unreamed femurs and nearly 22 for the reamedfemurs

6 Advances in Orthopedics

0 20 40 60 800

20

40

60

80

100

Voxe

l (

)

Defect aloneDefect and reamed

Time (days)

lowastlowastlowast

lowast

lowast

lowast lowast

Emptysoft tissue reaction (minus1000 to +800HU)

(a)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowast

Woven bone (+1200 to +1900 HU)

(b)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowastlowastlowastlowastlowast

lowastlowastlowast

lowastlowastlowast

Compact bone (gt+2700 HU)

(c)

Figure 4 Bone healing as a function of time for a defect without reaming and a defect with reaming (a) The results were quantified andplotted as a percentage of voxels representing nonorganized material as a function of time for unreamed femurs with a defect and reamedfemurs with a defect The percentages of (b) woven and (c) dense bone formation as a function of time were also graphed The results wereobtained from nine mice (119899 = 9 lowast119875 lt 005 lowastlowastlowast119875 lt 00001)

4 Discussion

120583CT is a precise tool for bone imaging and in recent decadesit has been shown to be the gold standard for studying bonestructure [30] It has been shown to be a relevant tool forcraniofacial studies in small rodents in which it is possibleto create critical size bone or nonhealing defects [8ndash11]Unfortunately such critical size defects cannot be created inthe long bones of mice without using nails or other materialsthat can impair 120583CT imaging [31 32] On the other handsmall noncritical size bone defects heal spontaneously andrapidly which may hinder the effectiveness of experimentaltreatments [16] To widen this ldquotherapeuticalrdquo window wereamed the endosteum of the medullar region of the bone

The aim of our study was to evaluate the utility of usinglongitudinal low dose 120583CT scans to assess the healing ofnoncritical size defects (1mm) in long bones in mice and toinvestigate the effect of intramedullary reaming on healingtimes

Longitudinal studies of small cortical defects using lowdose 120583CT scan are challenging for many reasons First bonedefects created using small drill bits are subject to operatorbias due to differences in the angle at which the drill bitenters the femur and due to micromovements once the bit isfully introduced This bias must be taken into account giventhe size of the defects (1mm) which are much smaller thancalvarial defects which can be up to 5mm in diameter inmice [8] Second small defects are assessed using multiple

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

Advances in Orthopedics 3

was extended proximally through the quadriceps to exposethe distal femur by lateral patellar eversion A 36410158401015840 drill bit(Model 865 Dremel Mount Prospect IL USA) was used tocreate a 11mm-diameter circular defect in the anterior cortexof both femurs 4mm above the femoral condylesThe defectswere irrigated with normal saline to remove bone residuesIntramedullary reaming was performed on one femur bydrilling a 13210158401015840-hole through the intercondylar notch andthen reaming the full length of the medullar cavity Thereamer was kept in the medullar cavity at high speed for 10 s

The defects were then covered by reclining the quadricepsback to its original position and repositioning the patellaThe incision was closed using nonabsorbable sutures Thesurgerywas performedby an orthopedic surgeonusing sterilematerials in aseptic conditions The mice were used forlive 120583CT imaging andor were euthanized for histologicalexaminations

23 Low Dose Microcomputed Tomography (120583CT) Scan Imag-ing Low dose 120583CT scan imaging was performed using aGamma Medica Triumph X-O small-animal CT scannercomposed of a 40W X-ray tube with a 75 120583m focal spotdiameter and a 2240 times 2368CsI flat panel X-ray detectorThe detector pixel size was 50 120583m and a 2 times 2 pixel binningscheme was used Scans were performed at 60 kVp and230 120583A using 512 projections in fly mode to reduce exposurewith a 5920mm FOV Images were reconstructed using thegeneral purpose reconstruction FBP kernel from the GammaMedica software over an isotropic 116120583m voxel spacinggrid providing a voxel volume of 156 times 10minus3mm3 Air andwater phantoms (15mL Falcon tubes) were scanned for eachimaging session to allow normalization in HU The live micewere scanned once prior to surgery as a control and thenat days 0 (immediately after surgery) 7 14 21 30 35 4457 and 78 after surgery To minimize movement the micewere isoflurane-sedated during the 120583CT imaging The doseof radiation administered was 117 cGy per scan as measuredby dosimeter

24 Software to Quantify Bone Healing Custom softwarewas written to automatically calculate the size of the cor-tical defects The 120583CT scans consisted of 512 individualimage slices stacked in a RAW file and converted to theNIFTI format using Fiji freeware (ImageJ Version 142qNational Institutes of Health USA) [20] ITK-SNAP freeware(httpwwwitksnaporg) was used to segment structures [21]and manually delineate 3D ROI on the cortical defects inthe postsurgery (day 0) scans Tracing an irregular anatomiccontour adjacent to the cortical surface was deemed the bestapproach to create the 3D ROI We used software providedby the Image Analysis and Visualization Platform (PAVISherbrookeQC) to virtually extract align and save images ofeachmouse femur to smaller volumesThe virtually extractedfemurs were then registered to the day 0 postsurgery refer-ence femur using FSL Flirt [22 23] Since we did not want thespatial information of the bones to be lost we parameterizedFlirt to use a rigid body transformation with 6 degreesof freedom After registration a linear transformation was

applied to the voxel signal intensities to convert them to theHU [24] scale using water for calibration This allowed thesignals from different scans made at different time pointsto be calibrated The HU scale makes it possible to conductanalyses of absolute voxel signal intensities in order todetermine the material density in each voxel (bone musclewater etc) We then calculated the kinetics of cortical defecthealing using the previously determined 3D ROI whichbecame relevant for each time point after registration Thevolumes and sizes of the cortical defects in treated and controlfemurs for each mouse were then quantified over time aswere their densities based on the HU scale The HU rangesfor empty or soft tissue reaction (eg hematoma) (minus1000 to+800HU) woven bone (+1200 to +1900HU) and compactbone (gt2700HU comparable to intact bone) were basedon published results from other groups [25ndash27] as well ason results from specific anatomical sites with known tissuecompositions

25 Statistics To calculate the sample size required we usedthe following repeated sample size estimation formula (119899 =2 + 119862(119904119889)

2) where 119904 is the standard deviation and 119889 is the

difference measured with a 119862 of 785 a power of 80 andan alpha of 005 The difference 119889 was measured between thevolumes of the cortical deficits in the left and right femurson postoperative day 29 when the dissimilarity between thetwo femurs was themostmarkedThe standard deviation wasalso calculated using the results on postoperative day 29 Thestandard deviation was 0255 and the measured differencewas 036The sample size required for a significant result wasthus at least 6 mice

One outcome of the study was the assessment of theagreement among three observers with respect to their3D ROI determinations for the same 120583CT scan This wasrepresented by a Bland-Altman plot [28] Graphically themean differences and the 95 confidence intervals (CI)were represented An intraclass coefficient (ICC) was alsoused to assess their reliability and validity The ICC for theinterobserver agreement was obtained by comparing blindmeasurements of the three observers For clinical studiesa Cronbach alpha value over 085 is considered very goodwhile a value over 09 is considered ldquoidealrdquo [29]

The main outcome of the present study was to assessthe differences in 3D ROI voxel HU cortical defect volumesbetween the reamed and unreamed femurs relative to therespective postoperative day 0 volumes The comparisonof means between reamed and unreamed samples at thesame time point was analyzed using two-way ANOVA andSidakrsquosmultiple comparisons test119875 values less than 005wereconsidered significant

Statistical analyses were performed using SPSS v2000and the results were graphed using Prism v50

3 Results

31 3D ROI Analysis of Interobserver Agreement Since initial3D ROI segmentation may be subjective and thus overesti-mate or underestimate healing we verified the interobserver

4 Advances in Orthopedics

Average (voxels)

500 1000 1500 20000

200

400

IInte

robs

erve

r diff

eren

ces (

voxe

ls)

Defect alone

A-BB-CC-ABias 007SD bias 1132

minus400

minus200

(a)

500 1000 1500 20000

200

400

Average (voxels)

Defect and reamed

Inte

robs

erve

r diff

eren

ces (

voxe

ls)

Bias minus007SD bias 1233

minus400

minus200

A-BB-CC-A

(b)

Figure 2 Interobserver agreement in the determination of 3DROIABland-Altmanplotwas used to graph differences in the absolute numberof voxels per 3D ROI as a function of the average size of the defect This made it possible to assess the agreement between three observers inthe determination of the 3D ROI Eighteen 120583CT images from (a) unreamed femurs with a defect and (b) reamed femurs with a defect wereblindly analyzed The two-by-two comparison of the three observers (A-B B-C and C-A) showed a 95 confidence interval from minus2265 to+2264 and from minus2459 to +2471 for unreamed femurs with a defect and reamed femurs with a defect respectively An interclass correlation(ICC) was used to compare interobserver variability The ICC scores were 0864 (very good 119875 lt 00001) and 0905 (ideal 119875 lt 00001) forthe unreamed femurs with a defect and the reamed femurs with a defect respectively

agreement which was represented using Bland-Altman plots(Figure 2) We used three observers (A-B-C) to blindlydetermine the 3DROI of the defectsThe results are presentedas the difference in the absolute number of voxels per3D ROI as a function of the average size of the defectOf the 36 scans 18 were of femurs with a defect but noreaming (Figure 2(a)) and 18 were of femurs with a defectand reaming (Figure 2(b)) The two-by-two comparison ofthe three observers (A-B B-C and C-A) showed that the 95confidence interval of the evaluations of the observers rangedfromminus2265 to +2264 (219plusmn 50) and fromminus2459 to +2471(201 plusmn 32) for defects without reaming and defects withreaming respectively

To determine whether the interobserver agreement wassignificant we calculated an interclass correlation (ICC)which allowed us to compare interobserver variability for agiven sampleThe ICC was 0864 (very good 119875 lt 00001) forthe defect without reaming and 0905 (ideal 119875 lt 00001) forthe defect with reaming indicating very good interobserveragreement In addition there was no significant differencebetween the 3D ROI of the defect without reaming and thedefect with reaming

32 Longitudinal Low Dose 120583CT Analysis of Bone HealingWe determined the difference in the kinetics of bone healingbetween reamed and unreamed femurs by calculating thenumber of voxels in nonorganized substrate woven boneand compact bone Figure 3(a) shows reconstructed images

of anteroposterior views of the femurs with their defects ondays 0 and 21 as well as on day 57 after surgery when themostsignificant variations in tissue composition of the reamed andunreamed femurs were observed In addition to visualizingthe registration of the femurs at different times it is possibleto identify the position of the 3D ROIs overlaid on the defectsalong different views (Figure 3(b)) Cross-sectional views ofthe femurs can also be seen in which the 3D ROI is apparent(Figure 3(c)) In both representations the voxels of the 3DROI are color-coded to differentiate between the variouscomponents in the defect that is emptysoft tissue reaction(red) woven bone (blue) and compact bone comparable tointact bone (green) Figures 3(b) and 3(c) show that wovenbone and compact bone formation is delayed in the reamedpreparation compared to the unreamed preparation

We used our custom software to verify the filling kineticsof the defects (Figure 4(a)) Graphing the percentage of voxelsin the minus1000 to +800HU range within the 3D ROI over timerevealed that reaming caused a significant delay in healingIt took 11 and 16 days after surgery for 50 healing to occurin the unreamed and reamed femurs respectively while 75healing took approximately 16 and 36 days respectively Byday 57 there was no significant difference in healing betweenthe two surgical procedures

We also assessed the formation of woven bone by graph-ing the percentage of voxels in the +1200 to +1900HU rangewithin the 3D ROI as a function of time (Figure 4(b)) Thepercentage of voxels related to woven bone was significantly

Advances in Orthopedics 5

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

2mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(a)

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

Axi

alC

oron

alSa

gitta

lA

xial

Cor

onal

Sagi

ttal

Axi

alC

oron

alSa

gitta

l

1mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(b)

Defectand reamedDefect alone

Day

0D

ay 2

1D

ay 5

7

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

1mm

(c)

Figure 3 Bone healing monitoring as a function of time using low dose 120583CT scans of live mice (a) Representative views of 3D 120583CTreconstructed images of femurs on days 0 21 and 57 after surgery Variations in the density of the 3D ROI (determined at day 0 after surgery)make it possible to estimate the tissue composition (b) 3D ROI can also be observed in axial coronal and sagittal views (c) Representativecross-sectional views of the femur in which the 3D ROI is apparent in one image (slice) The colored voxels make it possible to see changesin defect composition as a function of time where most of the voxels were red on day 0 and blue on day 28 with an increasing proportionof green on day 57 Nonorganized substrate (red minus1000 to +800HU) woven bone (blue +1200 to +1900HU) and compact bone (greengt2700HU)

higher in the unreamed preparations with the percent-age reaching a plateau around day 29 of 40 and 29for the unreamed and reamed preparations respectivelyInterestingly the percentages decreased to 20 by day 57when no significant difference between the preparations wasobserved

We investigated the formation of compact bone by graph-ing the percentage of voxels in the 3D ROI with gt2700HU

as a function of time (Figure 4(c)) The percentage of voxelsrelated to dense bone increased significantly faster in theunreamed femurs than in the reamed femurs This differencewas observed by day 35 when the percentage reached 30for the unreamed femurs but only 8 for the reamed femursThe percentage increased until day 57 reaching nearly 40for the unreamed femurs and nearly 22 for the reamedfemurs

6 Advances in Orthopedics

0 20 40 60 800

20

40

60

80

100

Voxe

l (

)

Defect aloneDefect and reamed

Time (days)

lowastlowastlowast

lowast

lowast

lowast lowast

Emptysoft tissue reaction (minus1000 to +800HU)

(a)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowast

Woven bone (+1200 to +1900 HU)

(b)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowastlowastlowastlowastlowast

lowastlowastlowast

lowastlowastlowast

Compact bone (gt+2700 HU)

(c)

Figure 4 Bone healing as a function of time for a defect without reaming and a defect with reaming (a) The results were quantified andplotted as a percentage of voxels representing nonorganized material as a function of time for unreamed femurs with a defect and reamedfemurs with a defect The percentages of (b) woven and (c) dense bone formation as a function of time were also graphed The results wereobtained from nine mice (119899 = 9 lowast119875 lt 005 lowastlowastlowast119875 lt 00001)

4 Discussion

120583CT is a precise tool for bone imaging and in recent decadesit has been shown to be the gold standard for studying bonestructure [30] It has been shown to be a relevant tool forcraniofacial studies in small rodents in which it is possibleto create critical size bone or nonhealing defects [8ndash11]Unfortunately such critical size defects cannot be created inthe long bones of mice without using nails or other materialsthat can impair 120583CT imaging [31 32] On the other handsmall noncritical size bone defects heal spontaneously andrapidly which may hinder the effectiveness of experimentaltreatments [16] To widen this ldquotherapeuticalrdquo window wereamed the endosteum of the medullar region of the bone

The aim of our study was to evaluate the utility of usinglongitudinal low dose 120583CT scans to assess the healing ofnoncritical size defects (1mm) in long bones in mice and toinvestigate the effect of intramedullary reaming on healingtimes

Longitudinal studies of small cortical defects using lowdose 120583CT scan are challenging for many reasons First bonedefects created using small drill bits are subject to operatorbias due to differences in the angle at which the drill bitenters the femur and due to micromovements once the bit isfully introduced This bias must be taken into account giventhe size of the defects (1mm) which are much smaller thancalvarial defects which can be up to 5mm in diameter inmice [8] Second small defects are assessed using multiple

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

4 Advances in Orthopedics

Average (voxels)

500 1000 1500 20000

200

400

IInte

robs

erve

r diff

eren

ces (

voxe

ls)

Defect alone

A-BB-CC-ABias 007SD bias 1132

minus400

minus200

(a)

500 1000 1500 20000

200

400

Average (voxels)

Defect and reamed

Inte

robs

erve

r diff

eren

ces (

voxe

ls)

Bias minus007SD bias 1233

minus400

minus200

A-BB-CC-A

(b)

Figure 2 Interobserver agreement in the determination of 3DROIABland-Altmanplotwas used to graph differences in the absolute numberof voxels per 3D ROI as a function of the average size of the defect This made it possible to assess the agreement between three observers inthe determination of the 3D ROI Eighteen 120583CT images from (a) unreamed femurs with a defect and (b) reamed femurs with a defect wereblindly analyzed The two-by-two comparison of the three observers (A-B B-C and C-A) showed a 95 confidence interval from minus2265 to+2264 and from minus2459 to +2471 for unreamed femurs with a defect and reamed femurs with a defect respectively An interclass correlation(ICC) was used to compare interobserver variability The ICC scores were 0864 (very good 119875 lt 00001) and 0905 (ideal 119875 lt 00001) forthe unreamed femurs with a defect and the reamed femurs with a defect respectively

agreement which was represented using Bland-Altman plots(Figure 2) We used three observers (A-B-C) to blindlydetermine the 3DROI of the defectsThe results are presentedas the difference in the absolute number of voxels per3D ROI as a function of the average size of the defectOf the 36 scans 18 were of femurs with a defect but noreaming (Figure 2(a)) and 18 were of femurs with a defectand reaming (Figure 2(b)) The two-by-two comparison ofthe three observers (A-B B-C and C-A) showed that the 95confidence interval of the evaluations of the observers rangedfromminus2265 to +2264 (219plusmn 50) and fromminus2459 to +2471(201 plusmn 32) for defects without reaming and defects withreaming respectively

To determine whether the interobserver agreement wassignificant we calculated an interclass correlation (ICC)which allowed us to compare interobserver variability for agiven sampleThe ICC was 0864 (very good 119875 lt 00001) forthe defect without reaming and 0905 (ideal 119875 lt 00001) forthe defect with reaming indicating very good interobserveragreement In addition there was no significant differencebetween the 3D ROI of the defect without reaming and thedefect with reaming

32 Longitudinal Low Dose 120583CT Analysis of Bone HealingWe determined the difference in the kinetics of bone healingbetween reamed and unreamed femurs by calculating thenumber of voxels in nonorganized substrate woven boneand compact bone Figure 3(a) shows reconstructed images

of anteroposterior views of the femurs with their defects ondays 0 and 21 as well as on day 57 after surgery when themostsignificant variations in tissue composition of the reamed andunreamed femurs were observed In addition to visualizingthe registration of the femurs at different times it is possibleto identify the position of the 3D ROIs overlaid on the defectsalong different views (Figure 3(b)) Cross-sectional views ofthe femurs can also be seen in which the 3D ROI is apparent(Figure 3(c)) In both representations the voxels of the 3DROI are color-coded to differentiate between the variouscomponents in the defect that is emptysoft tissue reaction(red) woven bone (blue) and compact bone comparable tointact bone (green) Figures 3(b) and 3(c) show that wovenbone and compact bone formation is delayed in the reamedpreparation compared to the unreamed preparation

We used our custom software to verify the filling kineticsof the defects (Figure 4(a)) Graphing the percentage of voxelsin the minus1000 to +800HU range within the 3D ROI over timerevealed that reaming caused a significant delay in healingIt took 11 and 16 days after surgery for 50 healing to occurin the unreamed and reamed femurs respectively while 75healing took approximately 16 and 36 days respectively Byday 57 there was no significant difference in healing betweenthe two surgical procedures

We also assessed the formation of woven bone by graph-ing the percentage of voxels in the +1200 to +1900HU rangewithin the 3D ROI as a function of time (Figure 4(b)) Thepercentage of voxels related to woven bone was significantly

Advances in Orthopedics 5

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

2mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(a)

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

Axi

alC

oron

alSa

gitta

lA

xial

Cor

onal

Sagi

ttal

Axi

alC

oron

alSa

gitta

l

1mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(b)

Defectand reamedDefect alone

Day

0D

ay 2

1D

ay 5

7

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

1mm

(c)

Figure 3 Bone healing monitoring as a function of time using low dose 120583CT scans of live mice (a) Representative views of 3D 120583CTreconstructed images of femurs on days 0 21 and 57 after surgery Variations in the density of the 3D ROI (determined at day 0 after surgery)make it possible to estimate the tissue composition (b) 3D ROI can also be observed in axial coronal and sagittal views (c) Representativecross-sectional views of the femur in which the 3D ROI is apparent in one image (slice) The colored voxels make it possible to see changesin defect composition as a function of time where most of the voxels were red on day 0 and blue on day 28 with an increasing proportionof green on day 57 Nonorganized substrate (red minus1000 to +800HU) woven bone (blue +1200 to +1900HU) and compact bone (greengt2700HU)

higher in the unreamed preparations with the percent-age reaching a plateau around day 29 of 40 and 29for the unreamed and reamed preparations respectivelyInterestingly the percentages decreased to 20 by day 57when no significant difference between the preparations wasobserved

We investigated the formation of compact bone by graph-ing the percentage of voxels in the 3D ROI with gt2700HU

as a function of time (Figure 4(c)) The percentage of voxelsrelated to dense bone increased significantly faster in theunreamed femurs than in the reamed femurs This differencewas observed by day 35 when the percentage reached 30for the unreamed femurs but only 8 for the reamed femursThe percentage increased until day 57 reaching nearly 40for the unreamed femurs and nearly 22 for the reamedfemurs

6 Advances in Orthopedics

0 20 40 60 800

20

40

60

80

100

Voxe

l (

)

Defect aloneDefect and reamed

Time (days)

lowastlowastlowast

lowast

lowast

lowast lowast

Emptysoft tissue reaction (minus1000 to +800HU)

(a)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowast

Woven bone (+1200 to +1900 HU)

(b)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowastlowastlowastlowastlowast

lowastlowastlowast

lowastlowastlowast

Compact bone (gt+2700 HU)

(c)

Figure 4 Bone healing as a function of time for a defect without reaming and a defect with reaming (a) The results were quantified andplotted as a percentage of voxels representing nonorganized material as a function of time for unreamed femurs with a defect and reamedfemurs with a defect The percentages of (b) woven and (c) dense bone formation as a function of time were also graphed The results wereobtained from nine mice (119899 = 9 lowast119875 lt 005 lowastlowastlowast119875 lt 00001)

4 Discussion

120583CT is a precise tool for bone imaging and in recent decadesit has been shown to be the gold standard for studying bonestructure [30] It has been shown to be a relevant tool forcraniofacial studies in small rodents in which it is possibleto create critical size bone or nonhealing defects [8ndash11]Unfortunately such critical size defects cannot be created inthe long bones of mice without using nails or other materialsthat can impair 120583CT imaging [31 32] On the other handsmall noncritical size bone defects heal spontaneously andrapidly which may hinder the effectiveness of experimentaltreatments [16] To widen this ldquotherapeuticalrdquo window wereamed the endosteum of the medullar region of the bone

The aim of our study was to evaluate the utility of usinglongitudinal low dose 120583CT scans to assess the healing ofnoncritical size defects (1mm) in long bones in mice and toinvestigate the effect of intramedullary reaming on healingtimes

Longitudinal studies of small cortical defects using lowdose 120583CT scan are challenging for many reasons First bonedefects created using small drill bits are subject to operatorbias due to differences in the angle at which the drill bitenters the femur and due to micromovements once the bit isfully introduced This bias must be taken into account giventhe size of the defects (1mm) which are much smaller thancalvarial defects which can be up to 5mm in diameter inmice [8] Second small defects are assessed using multiple

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

Advances in Orthopedics 5

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

2mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(a)

Day

0D

ay 2

1D

ay 5

7

Defectand reamedDefect alone

Axi

alC

oron

alSa

gitta

lA

xial

Cor

onal

Sagi

ttal

Axi

alC

oron

alSa

gitta

l

1mm

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

(b)

Defectand reamedDefect alone

Day

0D

ay 2

1D

ay 5

7

Emptysoft tissue reaction(minus1000 to +800)Woven bone (+1200 to +1900)Compact bone (gt2700)

ROI represented as HU density

1mm

(c)

Figure 3 Bone healing monitoring as a function of time using low dose 120583CT scans of live mice (a) Representative views of 3D 120583CTreconstructed images of femurs on days 0 21 and 57 after surgery Variations in the density of the 3D ROI (determined at day 0 after surgery)make it possible to estimate the tissue composition (b) 3D ROI can also be observed in axial coronal and sagittal views (c) Representativecross-sectional views of the femur in which the 3D ROI is apparent in one image (slice) The colored voxels make it possible to see changesin defect composition as a function of time where most of the voxels were red on day 0 and blue on day 28 with an increasing proportionof green on day 57 Nonorganized substrate (red minus1000 to +800HU) woven bone (blue +1200 to +1900HU) and compact bone (greengt2700HU)

higher in the unreamed preparations with the percent-age reaching a plateau around day 29 of 40 and 29for the unreamed and reamed preparations respectivelyInterestingly the percentages decreased to 20 by day 57when no significant difference between the preparations wasobserved

We investigated the formation of compact bone by graph-ing the percentage of voxels in the 3D ROI with gt2700HU

as a function of time (Figure 4(c)) The percentage of voxelsrelated to dense bone increased significantly faster in theunreamed femurs than in the reamed femurs This differencewas observed by day 35 when the percentage reached 30for the unreamed femurs but only 8 for the reamed femursThe percentage increased until day 57 reaching nearly 40for the unreamed femurs and nearly 22 for the reamedfemurs

6 Advances in Orthopedics

0 20 40 60 800

20

40

60

80

100

Voxe

l (

)

Defect aloneDefect and reamed

Time (days)

lowastlowastlowast

lowast

lowast

lowast lowast

Emptysoft tissue reaction (minus1000 to +800HU)

(a)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowast

Woven bone (+1200 to +1900 HU)

(b)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowastlowastlowastlowastlowast

lowastlowastlowast

lowastlowastlowast

Compact bone (gt+2700 HU)

(c)

Figure 4 Bone healing as a function of time for a defect without reaming and a defect with reaming (a) The results were quantified andplotted as a percentage of voxels representing nonorganized material as a function of time for unreamed femurs with a defect and reamedfemurs with a defect The percentages of (b) woven and (c) dense bone formation as a function of time were also graphed The results wereobtained from nine mice (119899 = 9 lowast119875 lt 005 lowastlowastlowast119875 lt 00001)

4 Discussion

120583CT is a precise tool for bone imaging and in recent decadesit has been shown to be the gold standard for studying bonestructure [30] It has been shown to be a relevant tool forcraniofacial studies in small rodents in which it is possibleto create critical size bone or nonhealing defects [8ndash11]Unfortunately such critical size defects cannot be created inthe long bones of mice without using nails or other materialsthat can impair 120583CT imaging [31 32] On the other handsmall noncritical size bone defects heal spontaneously andrapidly which may hinder the effectiveness of experimentaltreatments [16] To widen this ldquotherapeuticalrdquo window wereamed the endosteum of the medullar region of the bone

The aim of our study was to evaluate the utility of usinglongitudinal low dose 120583CT scans to assess the healing ofnoncritical size defects (1mm) in long bones in mice and toinvestigate the effect of intramedullary reaming on healingtimes

Longitudinal studies of small cortical defects using lowdose 120583CT scan are challenging for many reasons First bonedefects created using small drill bits are subject to operatorbias due to differences in the angle at which the drill bitenters the femur and due to micromovements once the bit isfully introduced This bias must be taken into account giventhe size of the defects (1mm) which are much smaller thancalvarial defects which can be up to 5mm in diameter inmice [8] Second small defects are assessed using multiple

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

6 Advances in Orthopedics

0 20 40 60 800

20

40

60

80

100

Voxe

l (

)

Defect aloneDefect and reamed

Time (days)

lowastlowastlowast

lowast

lowast

lowast lowast

Emptysoft tissue reaction (minus1000 to +800HU)

(a)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowast

Woven bone (+1200 to +1900 HU)

(b)

Defect aloneDefect and reamed

Voxe

l (

)

Time (days)0 20 40 60 80

0

20

40

6080

100

lowastlowastlowastlowastlowastlowast

lowastlowastlowast

lowastlowastlowast

Compact bone (gt+2700 HU)

(c)

Figure 4 Bone healing as a function of time for a defect without reaming and a defect with reaming (a) The results were quantified andplotted as a percentage of voxels representing nonorganized material as a function of time for unreamed femurs with a defect and reamedfemurs with a defect The percentages of (b) woven and (c) dense bone formation as a function of time were also graphed The results wereobtained from nine mice (119899 = 9 lowast119875 lt 005 lowastlowastlowast119875 lt 00001)

4 Discussion

120583CT is a precise tool for bone imaging and in recent decadesit has been shown to be the gold standard for studying bonestructure [30] It has been shown to be a relevant tool forcraniofacial studies in small rodents in which it is possibleto create critical size bone or nonhealing defects [8ndash11]Unfortunately such critical size defects cannot be created inthe long bones of mice without using nails or other materialsthat can impair 120583CT imaging [31 32] On the other handsmall noncritical size bone defects heal spontaneously andrapidly which may hinder the effectiveness of experimentaltreatments [16] To widen this ldquotherapeuticalrdquo window wereamed the endosteum of the medullar region of the bone

The aim of our study was to evaluate the utility of usinglongitudinal low dose 120583CT scans to assess the healing ofnoncritical size defects (1mm) in long bones in mice and toinvestigate the effect of intramedullary reaming on healingtimes

Longitudinal studies of small cortical defects using lowdose 120583CT scan are challenging for many reasons First bonedefects created using small drill bits are subject to operatorbias due to differences in the angle at which the drill bitenters the femur and due to micromovements once the bit isfully introduced This bias must be taken into account giventhe size of the defects (1mm) which are much smaller thancalvarial defects which can be up to 5mm in diameter inmice [8] Second small defects are assessed using multiple

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

Advances in Orthopedics 7

scans of the animals which while immobilized are notin exactly the same position which in turn increases therisk of miscalculations To circumvent these experimentalbiases the corresponding femur of each animal was firstrigorously aligned in space making it possible to exactlysuperpose the 3DROIThe 3DROIs at day 0 (after operation)were then set and their coordinates were applied to all timepoints The results of each time point were relativized tothe initial 3D ROI at day 0 (before and after operation)By applying this procedure we were able to assess callusformation and mineralization over time by using calibratedvoxels expressed in HU that were applied to different tissuecomposition categories that is empty or soft tissue reaction(eg hematoma) woven bone and mineralized compactbone More importantly the bone healing process over timedescribed in the present study matched that of previous 120583CTand histological studies [33ndash35]

In addition our results showed that the healing of corticaldefects is significantly affected by the endosteal reamingprocedure This was shown by the significant delay in defectfilling and the formation of woven and compact bone overtime In terms of compositional analyses Hayward et alsuggested that the use of contrast-enhanced 120583CT scans mayimprove the evaluation of nonmineralized (eg cartilage)and mineralized tissues in fracture calluses [36] From ourexperience we had no difficulty in distinguishing betweenthese tissues This may be due to the ranges of HU that weused to categorize the tissues and the use of the 3D ROI atday 0 (prior to surgery) as a reference Our software made itpossible to follow changes in the tissue density of a healingregion as small as 1mm3 Low dose 120583CT analyses may thusprovide a reasonable estimate of the tissue mineralization ofhealing bone based on HU

Our study also showed that appropriate analysis softwareand suitable image resolution limit observer variability Sinceour software relies on manual tracings of an initial volume ofinterest (3D ROI) on native 120583CT scans interobserver differ-ences may occur given that the exact delineation of corticaldefects depends partly on the judgment of the observersWe hypothesized that the greatest interobserver variability iscaused by the semisubjective delineation of the initial corticaldefect which is a bias that may be compounded Howeverour ICC interobserver analysis showed that this bias is smallfor experienced observers and does not affect the resultsAlternatives include global and local threshold segmentationbut since our 3DROI was relatively small and well delineatedwe judged that manual tracing segmentation would be themost convenient This finding also implied that dependingon the type of experiment investigators should choose theappropriate 120583CT resolution to maintain an acceptable levelof precision

While low dose 120583CT scans have limited resolution com-pared to high dose scans our study showed that the resolutionis sufficient and importantly low dose scans avoid the biolog-ical risk associated with repeated radiation exposure [37ndash40]Laperre et al reported that they observed no hematologicaltoxicity or radiotoxic effect on bone microarchitecture withtriple radiation doses of 434 cGy per scan [39] which

correspond to a cumulative dose of 130 cGy In our study thecumulative dose was under 120 cGy which is far less than200 cGy the dose at which the tumor risk increases in micethus making any radiation-induced effect on bone healingunlikely [41]

The creation of defects by drilling a hole in the bonecortex is a widely used and effective model for studyingbone healing [16 35] and it is well known that it leads tospontaneous healing [16 42] Many reasons could explain thehealing delay observed in the reamed preparations Previousstudies have shown that vascular compromise can induce asignificant healing delay or even nonunion in rodent fracturemodels [43 44] Since medullar vascularization accountsfor about two-thirds of total bone vascularization [45] itis reasonable to think that its disruption could increasehealing time In addition a loss of medullar substance andendosteum may also contribute to the delay in healing sincethey contain progenitor cells that contribute to bone healing[35 46] Other groups have used different approaches to slowdown spontaneous bone healing To assess how transplantedMSC contribute to healing cortical bone defects Gao etal sublethally irradiated the hindlimbs of mice in order toprevent endogenous MSC from participating in the corticalhealing process [17] Prolonging bone healing by medullarreaming may be a useful and simple procedure for limitingspontaneous healing and widening the window for studyingthe effect of compounds and exogenous stem cells

5 Conclusion

We showed that low dose 120583CT scans can be used to quantifyand characterize bone healing in live mice Using 3D ROIcoordinates determined immediately following surgery itwas possible to evaluate variations in tissue density withinthe 3D ROI and show that reaming delays the healing ofcortical defects inmiceTheproposed experimental approachreduces the number of animals needed provides quantitativelongitudinal results and is reproducible between differentobservers Bone healing prolongation combined with 120583CTimaging to study small bone defects in live mice thus showspotential as a promising tool for future preclinical researchon bone healing

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to Dr David Alcoloumbre DrOtman Sarrhini and Genevieve Drouin for their assistanceand Drs Maxime Descoteaux Matthieu Dumont and FelixC Morency from PAVI Lu-Zhao Di is the recipient of ascholarship from the Canadian Institutes of Health Research(CIHR) Elisabeth Leblanc is the recipient of scholarshipsfrom the Fondation pour la Recherche et lrsquoEnseignement

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

8 Advances in Orthopedics

en Orthopedie de Sherbrooke (FREOS) and CIHR Guil-laume Grenier holds a New Investigator Award from theFonds de Recherche du Quebec-Sante (FRQS) This workwas supported by grants from CIHR the National Scienceand Engineering Research Council of Canada (NSERC) theCanada Foundation for Innovation (CFI) and the Reseaude Therapie Cellulaire du FRQS (TheCell) The Centredrsquoimagerie moleculaire de Sherbrooke is part of the FRQS-funded CRCHUS

References

[1] A M Parfitt M K Drezner F H Glorieux et al ldquoBone histo-morphometry standardization of nomenclature symbols andunits Report of the ASBMRHistomorphometry NomenclatureCommitteerdquo Journal of Bone andMineral Research vol 2 no 6pp 595ndash610 1987

[2] M E Oetgen G AMerrell NW TroianoM CHorowitz andM A Kacena ldquoDevelopment of a femoral non-union model inthe mouserdquo Injury vol 39 no 10 pp 1119ndash1126 2008

[3] J F Griffith and H K Genant ldquoBone mass and architecturedetermination state of the artrdquoBest PracticeampResearch ClinicalEndocrinology amp Metabolism vol 22 no 5 pp 737ndash764 2008

[4] L A Feldkamp S A Goldstein A M Parfitt G Jesion andM Kleerekoper ldquoThe direct examination of three-dimensionalbone architecture in vitro by computed tomographyrdquo Journal ofBone and Mineral Research vol 4 no 1 pp 3ndash11 1989

[5] J H Kinney N E Lane and D L Haupt ldquoIn vivo three-dimensional microscopy of trabecular bonerdquo Journal of Boneand Mineral Research vol 10 no 2 pp 264ndash270 1995

[6] M L Bouxsein S K Boyd B A Christiansen R E GuldbergK J Jepsen and R Muller ldquoGuidelines for assessment of bonemicrostructure in rodents usingmicro-computed tomographyrdquoJournal of Bone and Mineral Research vol 25 no 7 pp 1468ndash1486 2010

[7] A C Jones C H Arns A P Sheppard D W HutmacherB K Milthorpe and M A Knackstedt ldquoAssessment of boneingrowth into porous biomaterials using MICRO-CTrdquo Bioma-terials vol 28 no 15 pp 2491ndash2504 2007

[8] C M Cowan Y-Y Shi O O Aalami et al ldquoAdipose-derivedadult stromal cells heal critical-size mouse calvarial defectsrdquoNature Biotechnology vol 22 no 5 pp 560ndash567 2004

[9] C M Cowan T Aghaloo Y-F Chou et al ldquoMicroCT evalua-tion of three-dimensional mineralization in response to BMP-2doses in vitro and in critical sized rat calvarial defectsrdquo TissueEngineering vol 13 no 3 pp 501ndash512 2007

[10] T Aghaloo C M Cowan Y-F Chou et al ldquoNell-1-inducedbone regeneration in calvarial defectsrdquoTheAmerican Journal ofPathology vol 169 no 3 pp 903ndash915 2006

[11] S-Y Park K-H Kim K-T Koo et al ldquoThe evaluation of thecorrelation between histomorphometric analysis and micro-computed tomography analysis in AdBMP-2 induced boneregeneration in rat calvarial defectsrdquo Journal of Periodontal andImplant Science vol 41 no 5 pp 218ndash226 2011

[12] L C Johnson R W Johnson S A Munoz G R MundyT E Peterson and J A Sterling ldquoLongitudinal live animalmicro-CT allows for quantitative analysis of tumor-inducedbone destructionrdquo Bone vol 48 no 1 pp 141ndash151 2011

[13] J H Waarsing J S Day and H Weinans ldquoAn improvedsegmentationmethod for in vivo 120583CT imagingrdquo Journal of Boneand Mineral Research vol 19 no 10 pp 1640ndash1650 2004

[14] J H Waarsing J S Day J C van Der Linden et al ldquoDetectingand tracking local changes in the tibiae of individual rats a novelmethod to analyse longitudinal in vivo micro-CT datardquo Bonevol 34 no 1 pp 163ndash169 2004

[15] V David N Laroche B Boudignon et al ldquoNoninvasive invivo monitoring of bone architecture alterations in hindlimb-unloaded female rats using novel three-dimensionalmicrocom-puted tomographyrdquo Journal of Bone and Mineral Research vol18 no 9 pp 1622ndash1631 2003

[16] J E Henderson C Gao and E J Harvey ldquoSkeletal phenotypingin rodents tissue isolation and manipulationrdquo in OsteoporosisResearch G Duque and K Watanabe Eds Springer LondonUK 2011

[17] C Gao J Seuntjens G N Kaufman et al ldquoMesenchymalstem cell transplantation to promote bone healingrdquo Journal ofOrthopaedic Research vol 30 no 8 pp 1183ndash1189 2012

[18] Y-X He G Zhang X-H Pan et al ldquoImpaired bone healingpattern inmice with ovariectomy-induced osteoporosis a drill-hole defect modelrdquo Bone vol 48 no 6 pp 1388ndash1400 2011

[19] M Nagashima A Sakai S Uchida S Tanaka M Tanaka andT Nakamura ldquoBisphosphonate (YM529) delays the repair ofcortical bone defect after drill-hole injury by reducing terminaldifferentiation of osteoblasts in the mouse femurrdquo Bone vol 36no 3 pp 502ndash511 2005

[20] M D Abramoff P J Magelhaes and S J Ram ldquoImageprocessing with imageJrdquo Biophotonics International vol 11 no7 pp 36ndash42 2004

[21] P A Yushkevich J Piven H C Hazlett et al ldquoUser-guided 3Dactive contour segmentation of anatomical structures signifi-cantly improved efficiency and reliabilityrdquo NeuroImage vol 31no 3 pp 1116ndash1128 2006

[22] D N Greve and B Fischl ldquoAccurate and robust brain imagealignment using boundary-based registrationrdquo NeuroImagevol 48 no 1 pp 63ndash72 2009

[23] M Jenkinson P Bannister M Brady and S Smith ldquoImprovedoptimization for the robust and accurate linear registration andmotion correction of brain imagesrdquo NeuroImage vol 17 no 2pp 825ndash841 2002

[24] G N Hounsfield ldquoComputed medical imagingrdquo Science vol210 no 4465 pp 22ndash28 1980

[25] A Katsumata A Hirukawa S Okumura et al ldquoEffects of imageartifacts on gray-value density in limited-volume cone-beamcomputerized tomographyrdquo Oral Surgery Oral Medicine OralPathology Oral Radiology and Endodontology vol 104 no 6pp 829ndash836 2007

[26] P Mah T E Reeves andW D McDavid ldquoDeriving Hounsfieldunits using grey levels in cone beam computed tomographyrdquoDentomaxillofacial Radiology vol 39 no 6 pp 323ndash335 2010

[27] A Donneys N S Nelson S S Deshpande et al ldquoQuantifyingmineralization using bone mineral density distribution in themandiblerdquo Journal of Craniofacial Surgery vol 23 no 5 pp1502ndash1506 2012

[28] J M Bland and D G Altman ldquoApplying the right statisticsanalyses of measurement studiesrdquo Ultrasound in Obstetrics ampGynecology vol 22 no 1 pp 85ndash93 2003

[29] J M Bland and D G Altman ldquoCronbachrsquos alphardquo The BritishMedical Journal vol 314 no 7080 article 572 1997

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

Advances in Orthopedics 9

[30] J L Tremoleda M Khalil L L Gompels M Wylezinska-Arridge T Vincent and W Gsell ldquoImaging technologies forpreclinical models of bone and joint disordersrdquo EJNMMIResearch vol 1 no 1 pp 1ndash14 2011

[31] B De Man J Nuyts P Dupont G Marchal and P SuetensldquoMetal streak artifacts in X-ray computed tomography asimulation studyrdquo IEEE Transactions onNuclear Science vol 46no 3 pp 464ndash472 1999

[32] S D Cook L P Patron S L Salkeld K E Smith B Whitingand R L Barrack ldquoCorrelation of computed tomography withhistology in the assessment of periprosthetic defect healingrdquoClinical Orthopaedics and Related Research vol 467 no 12 pp3213ndash3220 2009

[33] H Uusitalo J Rantakokko M Ahonen et al ldquoA metaphysealdefect model of the femur for studies of murine bone healingrdquoBone vol 28 no 4 pp 423ndash429 2001

[34] T M Campbell W T Wong and E J Mackie ldquoEstablishmentof a model of cortical bone repair in micerdquo Calcified TissueInternational vol 73 no 1 pp 49ndash55 2003

[35] L Monfoulet B Rabier O Chassande and J-C FricainldquoDrilled hole defects in mouse femur as models of intramem-branous cortical and cancellous bone regenerationrdquo CalcifiedTissue International vol 86 no 1 pp 72ndash81 2010

[36] L N M Hayward C M J De Bakker L C Gerstenfeld M WGrinstaff and E F Morgan ldquoAssessment of contrast-enhancedcomputed tomography for imaging of cartilage during fracturehealingrdquo Journal of Orthopaedic Research vol 31 no 4 pp 567ndash573 2013

[37] L Yu X Liu S Leng et al ldquoRadiation dose reduction incomputed tomography techniques and future perspectiverdquoImaging in Medicine vol 1 pp 65ndash84 2009

[38] R Duran-Struuck and R C Dysko ldquoPrinciples of bonemarrowtransplantation (BMT) providing optimal veterinary and hus-bandry care to irradiated mice in BMT studiesrdquo Journal of theAmerican Association for Laboratory Animal Science vol 48 no1 pp 11ndash22 2009

[39] K Laperre M Depypere N van Gastel et al ldquoDevelopmentof micro-CT protocols for in vivo follow-up of mouse bonearchitecture without major radiation side effectsrdquo Bone vol 49no 4 pp 613ndash622 2011

[40] R J Klinck G M Campbell and S K Boyd ldquoRadiation effectson bone architecture in mice and rats resulting from in vivomicro-computed tomography scanningrdquo Medical Engineeringand Physics vol 30 no 7 pp 888ndash895 2008

[41] K S Crump P Duport H Jiang N S Shilnikova D Krewskiand J M Zielinski ldquoA meta-analysis of evidence for hormesisin animal radiation carcinogenesis including a discussion ofpotential pitfalls in statistical analyses to detect hormesisrdquoJournal of Toxicology and Environmental Health Part B CriticalReviews vol 15 no 3 pp 210ndash231 2012

[42] C Bosch B Melsen and K Vargervik ldquoImportance of thecritical-size bone defect in testing bone-regeneratingmaterialsrdquoThe Journal of Craniofacial Surgery vol 9 no 4 pp 310ndash3161998

[43] K Hietaniemi J Peltonen and P Paavolainen ldquoAn experimen-tal model for non-union in ratsrdquo Injury vol 26 no 10 pp 681ndash686 1995

[44] C Lu T Miclau D Hu and R S Marcucio ldquoIschemia leads todelayed union during fracture healing a mouse modelrdquo Journalof Orthopaedic Research vol 25 no 1 pp 51ndash61 2007

[45] F W Rhinelander ldquoThevascular response of bone to internalfixationrdquo in The Science and Practice of Intramedullary NailingC C Edwards Ed pp 25ndash29 Lea amp Febiger Philadelphia PaUSA 1987

[46] C Colnot ldquoSkeletal cell fate decisions within periosteum andbone marrow during bone regenerationrdquo Journal of Bone andMineral Research vol 24 no 2 pp 274ndash282 2009

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Disease Markers

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom