2009 Accuracy and reliability of CBCT measurements used in the determination of facial indices in the Lab setup. JCMFS

of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Automobiles

Published:

Views: 39 | Pages: 7

Extension: PDF | Download: 0

Share
Description
2009 Accuracy and reliability of CBCT measurements used in the determination of facial indices in the Lab setup. JCMFS
Tags
Transcript
  This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institutionand sharing with colleagues.Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third partywebsites are prohibited.In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further informationregarding Elsevier’s archiving and manuscript policies areencouraged to visit:http://www.elsevier.com/copyright  Author's personal copy Accuracy and repeatability of cone-beam computed tomography (CBCT)measurements used in the determination of facial indicesin the laboratory setup * Bernard A. M. M. L. MOERENHOUT 1,4,5 , Frederik GELAUDE 2 , Gwen R. J. SWENNEN 1,4,5 ,Jan W. CASSELMAN 3,4,5 , Jos VAN DER SLOTEN 2 , Maurice Y. MOMMAERTS 1,4,51  Division of Maxillo-Facial Surgery, Department of Surgery, General Hospital St. Jan, Bruges, Belgium; 2  Division of Biomechanics and Engineering Design, Department of Mechanics, Katholieke Universiteit Leuven, Leuven, Belgium;  3  Department of Radiology and Medical Imaging, General Hospital St. Jan, Bruges, Belgium; 4  3-D Facial Imaging Research Group (3D-FIRG), General Hospital St. Jan, Bruges, Belgium;  5  Radboud University, Nijmegen, The Netherlands SUMMARY.  Aim:  To assess the three dimensional (3D) surface accuracy of a phantom’s face acquired fromacone-beam computed tomography(CBCT) scan and to determine the reliabilityof selected cephalometric mea-surements performed with Maxilim  software (Medicim N.V., Mechelen, Belgium).  Material and methods:  Amannequin head was imaged with a CBCT (I-CAT  , Imaging Sciences International, Inc., Hatfield, USA). Thedatawereusedtoproduce3Dsurfacemeshes(Maxilim  andMimics  ,MaterialiseN.V.,Leuven,Belgium)whichwere compared with an optical surface scan of the head using Focus Inspection  software (Metris N.V., Leuven,Belgium). The intra- and inter-observer reliability for the measurement of distances between facial landmarkswith Maxilim  3D cephalometry were determined by calculating Pearson correlation coefficients and intraclasscorrelation(ICC).The Dahlberg formulawasusedtoassessthemethoderror(ME).  Results:  (1) The maximalrange of the 3D mesh deviations was 1.9 mm for Maxilim  , and 1.8 mm for Mimics  segmentation. (2) Test e re-test and inter-observer reliability were high; Pearson’s correlation coefficient was 1.000 and the ICC was 0.9998.The ME of the vertical measurements was a little larger than that calculated for the width measurements. Max-imumME was 1.33 mm.  Conclusions:  The 3Dsurfaceaccuracyof CBCT scans segmentedwith Maxilim  andMimics  softwareis high. Maxilim  also shows satisfactory intra- and inter-assessor reliability for measurementof distances on a rigid facial surface.   2008 European Association for Cranio-Maxillofacial Surgery  Keywords:  face, facial index, radiography, cephalometry, surgery, maxillofacial, reproducibility of resultsINTRODUCTION One of the conclusions of a previous investigation on thedetermination of facial indices (  Mommaerts  and  Moer-enhout  , 2008) was that the measurements of the dis-tances, sellion e gnathion and supraorbitale e gnathion,with an anthropometric ruler were borderline reliable,since the minimum amount of surgical reduction in facialheight (maxillary impaction osteotomy, chin height re-duction osteotomy, closure of frontal open bite) is withinthe range of the measurement error, being 2 e 3 mm. Incontrast, inter-pupillary distance (IPD) measurement with a digital pupillometer was very reliable and moreprecise than the measurement of zygion e zygion withan anthropometric ruler.We suggested that the facial indices may be more ac-curately and reliably determined when using cone-beamcomputed tomography (CBCT) scans of seated patientswith lips relaxed and their lower jaw in rest position.The aim of this study was twofold. Firstly, to deter-mine the three dimensional (3D) surface accuracy of soft tissues acquired from a CBCT scan (I-CAT, ImagingSciences International, Hatfield, PA, USA) of a manne-quin head, and segmented using two commercial 3D im-age processing software (Maxilim  , Medicim N.V.,Mechelen, Belgium and Mimics  , Materialise N.V.,Leuven, Belgium), and secondly, assessing the reliabilityof distances between facial landmarks of the same phan-tom, as used for orthofacial surgery planning, computedby commercial 3D cephalometry software (Maxilim  ,Medicim N.V., Mechelen, Belgium). MATERIAL AND METHODS A polyester mannequin head painted with aqueouslacquer (P160.8100.0100.1104 Sporthead Liv weiß,Polyform GmbH & Co. KG, Rinteln, Germany) was con-sidered to provide the ideal facial surface for these tests;being immobile during the scanning procedure and un-changed between the scanning procedures (Fig. 1). It was scanned on three occasions, with 2-week intervals, * No grants, no company affiliation.18  Journal of Cranio-Maxillofacial Surgery  (2009)  37 , 18 e 23  2008 European Association for Cranio-Maxillofacial Surgerydoi:10.1016/j.jcms.2008.07.006, available online at  http://www.sciencedirect.com  Author's personal copy with an I-CAT Cone-Beam 3D Imaging System (ImagingSciences International, Hatfield, PA, USA  e  120 KVp;3 e 8 mA pulse mode; 0.4 mm typical voxel size; 2  20 sscan time; 17 cm diameter   22 cm height   e  extendedfield of view  e  scan dimensions). The I-CAT allows for the patient to be seated in an upright position, which isa prerequisite for determining linear and volumetric soft tissue dimensions in patients undergoing orthofacialreconstructive procedures. After the first scan (M0), themannequin head received gutta percha markers in drilledholesof0.5 mmdiameteratthefollowinglocations:supra-orbitale (left and right), sellion, zygion (left and right),pronasale,alare(leftandright),subnasale,labralesuperiusand inferius, stomion, menton and gnathion. The secondand third scans could be distinguished by the labels M1and M2 on the forehead, respectively.The mannequin head was transferred to the Depart-ment of Mechanical Engineering of the K.U. Leuvenfor acquisition of optical bone surface scans. More pre-cisely, the outside geometry of the mannequin headwas assessed using an optical measuring device (laser stripe scanning, LC50, Metris N.V., Leuven, Belgium).This device determines the distances between measure-ment points on the object and the scan head, and com-bines it with the 3D information from a coordinatemeasuring machine (MC16, Coord3 S.p.A, Bruzolo,Italy). The accuracy of the measuring system is  e  asspecified by the manufacturer   e  between 15 and35  m m. The measured point clouds were thinned/sampledusing a grid filter of size 0.025 mm, and converted intoa triangulated surface mesh of the mannequin’s face.The CBCT images (M0, M1 and M2) were stored us-ing DICOM 3.0 as a medical image file format (512 by512 pixels) into a Windows XP-based graphics worksta-tion (Pentium IV, 2.4 GHz, 512 MB RAM, NVIDIA Ge-Force4 Ti4400 graphics card) and subsequently either uploaded to the Medicim Company (Mechelen, Belgium)and imported into Maxilim  , version 2.0.1, 3D medicalimage processing software (Fig. 2); or sent to the Divi-sion of Biomechanics and Engineering Design (K.U.Leuven, Leuven, Belgium) and imported into theMimics  (Materialise N.V., Leuven, Belgium), version10.0, software. In both software packages, the samegrey value windowing was applied as segmentation. (i) Three dimensional surface accuracy of the face To assess the 3D surface accuracy of the phantom’s face,the segmented tissues of M1 and M2 were first convertedinto triangulated mesh representations using the March-ing Cubes algorithm (  Lorensen  and  Cline , 1987). For the Mimics approach, meshes were extracted using the‘optimal quality’ pre-defined mesh export setting.The obtained mesh data were compared with the opti-cal surface scan of the phantom. More precisely, the 3DEuclidian distance deviations from the optical surfacemesh to the segmented Maxilim   /Mimics  mesheswere assessed using the Focus Inspection  software(Metris N.V., Leuven, Belgium). The results of eachanalysis were available as distance colour maps, and assummarised statistical values. (ii) Reliability of distances between facial landmarks The .mxm file of the srcinal mannequin head (M0) wasused for Maxilim cephalometry by investigator BM onMay 10, May 24, June 7, June 22, July 10 and by inves-tigator MM on May 9, May 21, May 30, June 26 andJuly 9 of the year 2007. The primary parameters sellion,alare (left and right), endocanthion (left and right), pupil(left and right), subnasale, gnathion, zygion (left andright) and supraorbitale (left and right) were determinedusing original software setup and guidance ( Swennen et al., 2005). The secondary parameters (distances, no an-gles) were also computed by Maxilim  .Pearson correlation coefficients were performed to de-termine the reliability between the first and second mea-surements of MM (test  e retest reliability). The methoderror (ME) was calculated according to Dahlberg’s for-mula (  Dahlberg , 1940),  ME ¼  ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P d  2 = 2 n p   where  d   is Fig. 1 e The plaster mannequin head fixed in the I-CAT craniostat. Fig. 2 e M0 Maxilim scan of the mannequin head with the faciallandmarksusedforcomputationrepresentedasyellowdotsaftermanualdetermination.Accuracy and repeatability of CBCT 19  Author's personal copy the difference between the first and second measurementsand  n  is the number of comparisons performed.An intraclass correlation (ICC e two-way mixed effect model) was then calculated on the recorded measure-ments of both investigations MM and BM to determinethe level of inter-observer reliability.The statistical analyses were carried out with SPSS(Version 9.0, SPSS, Chicago, USA). RESULTS(i) Three dimensional surface accuracy of the face The DICOM files of the CBCT scan, reformatted byMaxilim  and Mimics  software, were compared withthe optical scan and proved to be an accurate representa-tion of the srcinal mannequin head. The range of the 3Dmesh deviations was 1.9 mm for M1 and M2 for Maxi-lim, and 1.8 mm for M1 and M2 for Mimics segmenta-tion (Figs. 3 and 4). Larger but circumscript errorswere seen at sharp transitions (eyelids, oral commisure,nasojugal grooves) and smaller ones (between   0.2and +0.2 mm) over larger surfaces in the nasal tip, cheek-bone and supraorbital areas (Figs. 5 and 6). Artefacts Fig. 3 e M2 Maxilim scan versus optical Metris scan, focussed onmaximum positive errors. Fig. 4 e M2 Maxilim scan versus optical Metris scan, focussed onmaximum negative errors. Fig. 5 e M2 Mimics scan versus optical Metris scan, generalcomparison with colour indication of error magnitude. Fig. 6 e Pie chart representing percentage of error magnitude asillustrated in Fig. 5.20  Journal of Cranio-Maxillofacial Surgery  Author's personal copy were visible in the forehead, where metal particles of 5 mm diameter were found in the polyester structure. (ii) Reliability of distances between facial landmarks,as measured with Maxilim  cephalometry The secondary parameters (distances, angles) of the re-peated cephalometric analyses are tabulated in Table 1.The D-MEs per measurement for investigators BM andMM are displayed in Table 2. Pearson’s  r   (test  e retest reliability) was 1.000 (significant at the 0.01 level  e two-tailed). The average measure ICC (inter-observer reliability) was 0.9998 (95 %  confidence interval (CI),lower  ¼ 0.9995, upper  ¼ 0.9999).The transverse dimension is very reliably determinedby the IPD (a measurement error of 0.08 and 0.21 mmfor BM and MM, respectively). Facial width measure-ment is just a little less reliable (ME of 0.24 and0.51 mm for BM and MM, respectively).There is not much difference in the ME of the measure-ments of the vertical dimension (sellion e gnathion, supra-orbitale left  e gnathion, supraorbitale right  e gnathion).The absolute ME is a little larger than that calculatedfor the width measurements, but facial height is alsogreater. DISCUSSION Computer tomography (CT) has brought to the craniofa-cial surgeon a 3D representation of anatomic structures.Multi-detector CT (MDCT) machines are commonly ac-cepted as the reference standard against which other de-vices have to compete in terms of geometric accuracy(  Mischkowski , 2007).  Richtsmeier   et al. (1995) concludedfrom their study that the average error in positioning an-atomic landmarks in three dimensions on CT slice imagesis always less than 0.5 mm. Although data collected fromCT scans were internally consistent, validation results Table 1 e The mean and standard deviation of the four repetitive cephalometries performed by investigator BM and MM on the M0 scanDistances between landmarks Investigator BM Investigator MM  Horizontal  Zygion left - zygion right 141.67 (0.21) 140.85 (0.59)Endocatnion left  e endocanthio right 39.85 (0.3) 38.85 (0.25)Alare left  e alare right 33.45 (0.19) 33.22 (0.49)Subnasale e gnathion 69.67 (1.42) 69 (1.22)Pupil left  e pupil right 67.3 (0.12) 67.47 (0.22) Vertical  Sellion e gnathion 119.5 (0.55) 117.05 (1.21)Supraorbitale right  e gnathion 120.35 (1.11) 119.67 (0.86)Supraorbitale left  e gnathion 120.2 (0.77) 120.17 (1.4)Ratios between linear measurementsZygion left-zygion right/supraorbital right  e gnathion 117.75 (1.18) 117.7 (0.82)Zygion left-zygion right/supraorbitale left  e gnathion 117.9 (0.81) 117.22 (1.28)Zygion left-zygion right/sellion-gnathion 118.57 (0.63) 120.32 (0.96)Endocanthion left  e endocanthion right/alare left  e alare right 119.05 (0.95) 117 (2.27)Subnasale-gnathion/sellion-gnathion 58.3 (1.39) 58.95 (0.47)Subnasale-gnathion/supraorbitale right  e gnathion 57.9 (1.63) 57.625 (0.7)Sunsale-gnathion/supraorbitale lieft  e gnathion 57.97 (1.53) 57.25 (0.17)Pupil right  e pupil left/supraorbitale left  e gnathion 56 (0.29) 56.17 (0.82)Pupil right  e pu left/supraorbitale right  e gnathion 55.92 (0.43) 56.37 (0.62) Table 2 e ME (method error acc. To Dahlberg), in millimetre for the distances, in percentage for the ratiosBM MM  Distances Zygion left  e zygion right 0.24 0.51Endocanthion left  e endocanthion right 0.24 0.20Alare left  e alare right 0.18 0.39Subnasale e gnathion 1.02 0.91Pupil right  e pupil left 0.08 0.21Sellion e gnathion 0.47 0.73Supraorbitale right  e gnathion 0.90 0.74Supraorbitale left  e gnathion 0.56 1.17  Ratios Zygion left  e zygion right/supraorbitale right  e gnathion 1.05 0.90Zygion left  e zygion right/supraorbitale left  e gnathion 0.69 1.33Zygion left  e zygion right/sellion e gnathion 0.60 0.81Endocanthion left  e endocanthion right/alare left  e alare right 1.07 1.74Subnasale e gnathion/sellion e gnathion 1.02 0.47Subnasale e gnathion/supraorbitale right  e gnathion 1.20 0.56Subnasale e gnathion/supraorbitale left  e gnathion 1.10 0.14Pupil right  e pupil left/supraobitale left  e gnathion 0.21 0.70Pupil right  e pu left/supraorbitale right  e gnathion 0.35 0.54Accuracy and repeatability of CBCT 21
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x