Evaluation of the accuracy of Cone Beam Computerized Tomography (CBCT): medical imaging technology in head and neck reconstruction
© Logan et al.; licensee BioMed Central Ltd. 2013
Received: 10 January 2013
Accepted: 24 February 2013
Published: 21 March 2013
With the introduction, development and commercialization of Cone Beam Computerized Tomography (CBCT) technologies in the field of head and neck reconstruction, clinicians now have increased access to the technology. Given the growth of this new user group, there is an increasing concern regarding proper use, understanding, quality and patient safety.
The present study was carried out to evaluate data acquisition of CBCT medical imaging technology and the accuracy of the scanning at three different machine warming times. The study also compared the accuracy of CBCT at 0.2 mm slice thickness and Computerized Tomography (CT) at 1 mm slice thickness. A control model was CT scanned at five random intervals, at 1 mm slice thickness and CBCT scanned at specialized intervals, at 0.2 mm slice thickness. The data was then converted and imported into a software program where a digital registration procedure was used to compare the average deviations of the scanned models to the control.
The study found that there was no statistically significant difference amongst the three CBCT machine warming times. There was a statistically significant difference between CT scanning with 1 mm slice thickness and CBCT scanning with 0.2 mm slice thickness.
The accuracy of the i-CAT CBCT scans used in the present study with a parameter at voxel size 0.2, will remain consistent and reliable at any warming stage. Also the difference between the CBCT i-CAT scans and the CT scans was not clinically significant based on suggested requirements of clinicians in head and neck reconstruction.
Three-dimensional (3D) information has become an important tool in assisting diagnosis and surgical planning in the oral and maxillofacial field. Many professionals in the field have been limited in the use of medical imaging technologies that produce 3D images and models due to availability, restricted access and radiation dose considerations . With the introduction, development and commercialization of CBCT technologies, clinicians now have increased access to the technology. Given the growth of this new user group, there is an increasing concern regarding proper use, adequate understanding, optimal quality and patient safety . It is important for users of this technology to be well informed and trained in all the technical implications of creating a proper and accurate 3D surface model. The quality of the CBCT scan depends on the scanner type, the scanning parameters and the reconstruction settings . There is a significant volume of information in the literature that describes the applications of CBCT technologies [3–7], but few studies examine the important technical protocols that need to be followed.
The present study is intended to contribute to the growing field of CBCT imaging technologies, contribute to the improvement of scanning techniques and protocols and bring knowledge to clinical users. Its general purpose is to examine the accuracy of CBCT and CT scanning technologies using 3D surface model reconstructions. There are two specific objectives. The first is to assess whether there is a difference in accuracy between CBCT scanning at three difference machine warming stages. The warming period of the CBCT was explored in order to understand whether sensitivity and precision of the sensor in the CBCT is affected by thermal distortion. The second is to assess whether there is a difference in accuracy between CT scanning and CBCT scanning at the three warming stages.
The milled model was then scanned using the Kreon laser scanner (Kreon Technologies,Limoge, France) mounted on a FARO Arm Laser Scanner (FARO Technologies Inc. Lake Mary, FL) in order to produce a representation of the model that was highly accurate. The FARO arm is a mechanical arm with a precision of 15 microns on which the Kreon 3D scanner is mounted that was precise to 40 microns. The combined precision of the scanner and the arm was 55 microns. The laser scan surface model produced by the 3D scanner was the control model for comparison.
CBCT Parameters (i-Cat 17–19, Imaging Sciences International LLC, Hatfield, PA USA) used for all scanning
Tube Voltage (kilovolt kV)
Tube Current (Milli-ampere mA)
Tube Current x Exposure Time
(Milli-ampere x seconds mAs)
CT Parameters (Toshiba Medical Systems, Aquilion 64) used for all scanning
1 mm slices
Anatomical plane data required
Occlusal plane position
Perpendicular to the gantry
Each CBCT scan was performed at the Institute for Reconstructive Sciences in Medicine (iRSM), Edmonton, Alberta, Canada. A voxel size of 0.2 mm was chosen in order to obtain the highest resolution for 3D volume reconstruction. All CT scans were completed by the lead CT technologist in the Department of Diagnostic Imaging (DI) at the Misericordia Community Hospital in Edmonton, Alberta, Canada. The CT protocol, which included 1 mm slices (using 1 mm overlap), was intended to mimic the guidelines of medical modeling used for patients of iRSM.
Threshold levels represented in Figure 3
Threshold levels represented in Figure 5
Once the optimal threshold was found, a consistent threshold was kept for all CBCT scans and for all CT scans. The resulting 3D models were then exported as stl files.
Accuracy of CBCT at three machine warming stages
Descriptive statistics for absolute average deviation (in mm) from the target control model
CBCT 1 hour
CBCT 4 hour
Accuracy of CT and CBCT at three machine warming stages
The present study was faced with a number of unanticipated issues that added complexity. Thresholding became an important technical challenge in the process of data acquisition and conversion in the study. The researcher was left to estimate the optimal threshold level and this affected the shape and accuracy of the 3D surface model reconstruction. Throughout the process of the study, the topic of thresholding was considered within the scope of the research project and discussed with an expert in the field of computer science as well as with technical support employees of the i-CAT imaging technology and software. Interestingly, the present study revealed ambiguity, inconsistency and a lack of understanding among several sources on this topic. It was decided to maintain consistency wherever possible in the thresholding procedure. Although a consistent threshold level was used in the present study due to limitations in the scope of the research project, it is important to note that every scan completed by the CBCT will have a different optimal threshold. For example, it was found that the optimal threshold for the CBCT scans completed on day five were all different. The optimal threshold for CBCT scan cold was −8, 1 hour was −20 and 4 hour was −21.
In general, the present study revealed important findings that contribute to the growing body of knowledge and technical understanding in the area of CBCT and medical imaging technology. In terms of its first objective, evaluating the accuracy of CBCT at three machine warming stages, the study showed no statistically significant differences among the periods for the machine tested. It is important to consider that the overall averages of the average deviation of all the warming periods were smaller than 0.005 of a mm (that is one 200th of a mm). This number possibly could have been smaller with optimal thresholding levels for each single scan. Given that the resulting averages were of very small magnitude, the clinical significance of the mean differences is considered to be of no concern. It was felt that clinicians can be confident that the accuracy of the i-CAT CBCT scans used in the present study with a parameter at voxel size 0.2, will remain consistent and reliable at any warming stage of the i-CAT CBCT machine. The result of the first objective is a positive outcome for clinicians as they can remain confident of the high accuracy (within 0.007 mm) of the different machine warming stages of the CBCT scanning technology tested.
In terms of the second objective, the study showed that there was a statistically significant difference in average deviation between CT scanning with 1 mm slice thickness and CBCT scanning with 0.2 mm slice thickness. CT scanning with 1 mm slice thickness resulted in an average deviation from the control of 0.056 mm, and was shown to be less accurate than the CBCT scanning with 0.2 mm slice thickness at all machine warming stages where an average deviation from the control of 0.007 mm was found. Although there was a statistically significant difference between CBCT at 0.2 mm slice thickness and CT at 1 mm slice thickness, it appeared that there was no clinical significance to the result. The researcher discussed the results of the study with a head and neck surgeon and with a maxillofacial prosthodontist. Both discussions revealed that the results of the study were not considered clinically significant in the respondents’ practice of surgical design and preoperative planning. The head and neck surgeon revealed that any deviation in bone position under 1 mm would not be considered clinically significant in terms of osseous reconstruction, while the maxillofacial prosthodontist revealed that any deviation under 0.5 mm would not be considered clinically significant in terms of implant installation planning. While the relevance of dimension to different clinical specialists needs scientific definition, the differences cited are both well beyond the average deviation from the control of 0.056 mm for CT and 0.007 mm for CBCT. The result of the second objective is also a positive outcome for clinicians as CT medical imaging technology is a common means by which data are obtained from their patients. These findings are specific to the CBCT machine, materials and techniques used in the present study.
The goal of the study on the accuracy of CBCT medical imaging technology was to assess whether there are differences between CT and CBCT in the form of 3D surface model reconstructions. For clinicians it is important to understand the parameters of the medical imaging technologies and what the differences are at different levels in order to obtain the optimal 3D surface model reconstruction.
Heather Logan (HL) - MSc, BDes
HL is a Surgical Design Simulationist at the Institute for Reconstructive Sciences in Medicine (iRSM). She completed her Master of Science in Rehabilitation Science with a focus in Surgical Design and Simulation in 2011 at the University of Alberta. HL divides her time between clinical work in facial prosthetics, surgical design and simulation and research. Her research involves developing and refining surgical design methods and developing digital processes for facial prosthetic fabrication.
Dr. Johan Wolfaardt - BDS, MDent (Prosthodontics), PhD
JW is a Director of the Institute of Reconstructive Sciences in Medicine (iRSM) and is appointed as a Full Professor in the Faculty of Medicine and Dentistry, University of Alberta, Canada. His clinical and research interests are in the area of maxillofacial prosthetics with particular emphasis in the area of head and neck reconstruction, osseointegration and treatment outcomes. JW has led the development of the research program at COMPRU. His research interests involve treatment outcomes, digital technologies in head and neck reconstruction and biomechanics of osseointegrated implants. JW has published over 80 papers in refereed journals and contributed to a variety of texts. He has lectured both nationally and internationally on maxillofacial prosthetics, osseointegration in head and neck reconstruction, challenges of introduction of advanced digital technology, knowledge work, teamwork and quality management. JW is elected to the Boards of the International Society of Maxillofacial Rehabilitation, the American Academy of Maxillofacial Prosthetics and the International College of Prosthodontists.
Dr. Hadi Seikaly - MD, FR CSC
HS is a Professor of Surgery in the Department of Surgery (University of Alberta), Director of the Division of Otolaryngology Head and Neck Surgery, and the Zone Section Head for Otolaryngology Head and Neck Surgery. In addition, he is presently the president of the University Hospital Medical Staff Society, serves on several local, national, and international administrative committees and is on the Executive Council of the Canadian Society of Otolaryngology. HS graduated from the University of Toronto medical school and completed his residency training at the University of Alberta in Otolaryngology Head and Neck Surgery. He then obtained fellowship training at the University of Texas Medical Branch in advanced head & neck oncology, microvascular reconstruction and facial cosmetic surgery. Dr. Seikaly returned to the University of Alberta as an attending in the division of Otolaryngology, department of surgery in 1996 where he has been active in teaching, patient care, and research.
HS continues to have a large practice dedicated to head, neck and skull base oncology and reconstruction. His research interests include functional surgical and reconstructive outcomes, microvascular head and neck reconstruction, submandibular gland transfer and medical modeling as it applies to the head and neck region.
He has published more than 50 papers in peer reviewed journals and numerous chapters.
Dr. Pierre Boulanger – Ph.D., P.Eng
PB worked for 18 years at the National Research Council of Canada as a senior research officer where his primary research interests were in 3D computer vision, rapid product development, and virtualized reality systems. He now holds a double appointment as a professor at the University of Alberta in the Department of Computing Science and in the Department of Radiology and Diagnostic Imaging (Faculty of Medicine). PB is the Director of the Advanced Man Machine Interface Laboratory as well as the scientific director of the Alberta Radiological Visualization Center. His main research topics and teachings are on virtualized reality systems and medical imaging. He is also a Principal Investigator for Stereo IPTV at TRLabs. PB has published more than 260 scientific papers in various journals and conferences. PB is on the editorial board of two major academic journals as well as on many international committees and frequently gives lectures on rapid product development and virtualized reality. On the commercial side, PB is the President of PROTEUS Consulting Inc., an Alberta-based consulting firm specialized in Virtual Reality Applications.
Dr. Bill Hodgetts – Ph.D.
BH obtained his B.A. in Psychology and his M.Sc. in Audiology at the University of Western Ontario. He received his Ph.D. in Rehabilitation Sciences at the University of Alberta. BH is an Associate Professor in the Department of Speech Pathology and Audiology at the University of Alberta where he teaches in the areas of hearing science/audiology and research methods and statistics. He has a joint appointment with the Institute for Reconstructive Sciences in Medicine (iRSM), where he is program director of Bone Conduction Amplification. BH’s research involves developing and refining the selection, verification, and validation of fitting procedures for BAHA (bone anchored hearing aid). He also has a research interest in bone anchored hearing aids and noise exposure from MP3 Players.
Cone beam computerized tomography
Computer aided design
Institute for reconstructive sciences in medicine
Digital imaging and communications in medicine
I wish to acknowledge the Covenant Health Research Centre Advisory Committee and the Institute for Reconstructive Sciences in Medicine at the Misericordia Community Hospital, Edmonton, AB for the funding to support this research project. I am thankful to Diagnostic Imaging at the Misericordia Community Hospital, Edmonton, AB for the medical imaging support. I thank Southern Implants (South Africa), Dr. Keith Aronyk and Fari Karimi-Boushehri for their help in the production of the control model for the CBCT study.
- Scarfe WC, Farman AG, Sukovic P: Clinical applications of cone-beam computed tomography in dental practice. J Can Dent Assoc. 2006, 72 (1): 75-80.PubMed
- De Vos W, Casselman J, Swennen GRJ: Cone-beam Computerized Tomography (CBCT) imaging of the oral and maxillofacial region: A systematic review of the literature. Int J Oral Maxillofac Surg. 2009, 38 (6): 609-625. 10.1016/j.ijom.2009.02.028.PubMedView Article
- Ito K, Gomi Y, Sato S, Arai Y, Shinoda K: Clinical application of a new compact CT system to assess 3-D images for the preoperative treatment planning of implants in the posterior mandible: a case report. Clin Oral Implants Res. 2001, 12 (5): 539-542. 10.1034/j.1600-0501.2001.120516.x.PubMedView Article
- Hashimoto K, Arai Y, Iwai K, Araki M, Kawashima S, Terakado M: A comparison of a new limited cone beam computed tomography machine for dental use with a multidetector row helical CT machine. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2003, 3;95 (3): 371-377.View Article
- Farman AG, Scarfe WC: The basics of maxillofacial Cone Beam Computed Tomography. Semin Orthod. 2009, 15 (1): 2-13. 10.1053/j.sodo.2008.09.001.View Article
- Sukovic P: Cone Beam Computed Tomography in craniofacial imaging. Orthod Craniofac Res. 2003, 6 (Suppl 1): 31-36. discussion 179PubMedView Article
- Ziegler CM, Woertche R, Brief J, Hassfeld S: Clinical indications for digital volume tomography in oral and maxillofacial surgery. Dentomaxillofacial Radiology. 2002, 31 (2): 126-130. 10.1038/sj.dmfr.4600680.PubMedView Article
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.