Dentomaxillofacial Radiology (2007) 36, 192197 q 2007 The British Institute of Radiology http:/ /dmfr.birjournals.

org

RESEARCH

Cone beam CT and conventional tomography for the detection of morphological temporomandibular joint changes
H Hintze*,1, M Wiese1,2 and A Wenzel1
1 Department of Oral Radiology, School of Dentistry, University of Aarhus, Aarhus, Denmark; 2Department of Radiology, School of Dentistry, University of Copenhagen, Copenhagen, Denmark

Objective: To compare the diagnostic accuracy of cone beam CT images with conventional tomographic images for the detection of morphological temporomandibular joint (TMJ) changes. Methods: 80 dry human skulls were scanned using a NewTomw 3G scanner and lateral and frontal reconstructions of the right and the left TMJs were performed. In addition, lateral and frontal crosssectional tomograms of the skulls TMJs were obtained in a Cranex Tome unit with Digora storage phosphor plates. Naked-eye inspection of the TMJs performed by three observers served as the gold standard for the true presence of morphological changes. The mandibular fossae were excluded from the study due to few changes in this joint component. The NewTomw and the conventional tomographic images were examined by three independent observers using a binary scale for the presence of morphological changes in the condyle (attening, defects and osteophytes) and the articular tubercle (attening and defects). The accuracy for the different types of changes in relation to the condyles and the articular tubercles was expressed as sensitivity and specicity values, whereas the diagnostic accuracy for a general assessment including all changes in both joint components was expressed by the sum of cases where the gold standard and the radiographic scores were not identical (absolute difference). Differences between the two radiographic modalities were tested by paired t-test. Results: Detection of the various types of morphological changes in relation to the condyle and the articular tubercle assessed separately resulted in no signicant differences between the two radiographic modalities, with the exception of bone defects in the articular tubercle examined on frontal views alone where the specicity with tomography was signicantly higher than with cone beam CT. Detection of all morphological changes in relation to both the condyle and the articular tubercle showed a signicantly higher accuracy with tomography than with cone beam CT using lateral views alone, but there was no signicant difference between the two modalities using frontal views alone and lateral and frontal views in combination. Conclusion: In general, no signicant differences in diagnostic accuracy for the detection of bone changes in the condyle and in the articular tubercle were found between cone beam CT images and conventional tomograms. Dentomaxillofacial Radiology (2007) 36, 192197. doi: 10.1259/dmfr/25523853 Keywords: radiography; temporomandibular joint (TMJ); digital radiography; cone beam CT; NewTom Introduction Recently, cone beam CT (CBCT) for dental and maxillofacial diagnostic osseous tasks has been developed as an alternative to conventional CT. The CBCT technology results in images
*Correspondence to: Hanne Hintze, Department of Oral Radiology, School of Dentistry, Faculty of Health Sciences, University of Aarhus, Vennelyst Boulevard 9, DK-8000 Aarhus C, Denmark; E-mail: hhintze@odont.au.dk Received 9 February 2006; revised 30 June 2006; accepted 3 July 2006

of CT-like quality obtained on the basis of less expensive equipment and components, shorter patient examination time and much lower radiation dose than required for conventional CT.1 5 In addition, the scanning procedure of the patient and the software for image reconstruction connected with CBCT are very user-friendly. During the past few years, CBCT systems have been introduced in many dental institutions (schools, centres,

TMJ radiography H Hintze et al

193

special clinics etc.). Here, the CBCT technology serves as a supplement to or a substitute for the traditional radiographic methods. Since the effective radiation dose with CBCT is still higher than with many of the traditional radiographic methods, such as full-mouth series and panoramic and skull views, CBCT should substitute for such examinations only in case its superiority (higher diagnostic accuracy, higher reproducibility etc.) outweighs its increased potential biological radiation risk to the patient. To assess when CBCT is preferable in dental patients, it is important to compare the accuracy of CBCT for all relevant diagnostic tasks with the traditionally applied methods. So far, CBCT performed with the NewTomw (NIM, Italy) and the Accuitomo (Morita Co., Japan) scanners has been found to be superior to panoramic radiography for the depiction of the inferior alveolar nerve6 and for the identication of articially made bone defects in pigs and dry human mandibles.7 In addition, CBCT performed with one of the above-mentioned scanners has been claimed to be effective for implant planning,8 10 postoperative assessment of zygomaticomaxillary fractures,11 localization of maxillary unerupted canines,12 assessment of the thickness of the roof of the glenoid fossa13 and examination of the temporomandibular joint (TMJ),14 but the methods diagnostic accuracy in comparison with traditionally applied methods such as linear and spiral tomography and perhaps stereo-scanography, has not been evaluated. The aim of the present study was to compare the diagnostic accuracy of CBCT performed with the NewTomw 3G scanner with conventional tomography obtained with the Cranex Tome unit for the detection of morphological bone changes in the TMJ in dry human skulls, from which the true presence of changes could be validated by direct visual inspection. Materials and methods The left and right TMJs in 80 dry human skulls were scanned in a NewTomw 3G CBCT scanner (NIM s.r.l., Verona, Italy) using the 9 inch detector eld and automatic exposure parameters depending on bone volume and density. The scanner operates at 110 kV (constant), 0.5 mA and an exposure time which varied between 5 s and 7 s. Each skull was positioned on a cast pillow placed in the scanners gantry. The skulls mid-sagittal plane was aligned with the scanners medial light beam, whereas the lateral light spot was centred at the level of the left condyle. Cotton rolls were placed between the skulls maxillary and mandibular teeth or the edentulous jaws to separate the condyle from direct contact with the mandibular fossa. A radiographer, a radiologist and two dental students under supervision performed the scans. The primary reconstruction of the raw data was restricted to the TMJ region (approximately 1 cm superior to the mandibular fossa and 1 cm inferior to the condylar neck), and a series of axial views of 1 mm thickness were automatically generated. On one of those axial views the long axis of the examined

condyle was traced, and the software generated lateral and frontal cross-sectional reconstructions perpendicular and parallel to the long axis of the condyle, respectively. The thickness of the image slices was 1 mm and the distance between slices was 0.5 mm for both lateral and frontal reconstructions. Conventional spiral tomography was performed as individually corrected (based on a lateral four angle preexamination) lateral (image plane perpendicular to the long axis of the condyle) and frontal (image plane parallel to the long axis of the condyle) tomograms in a Cranex Tome X-ray unit (Soredex, Helsinki, Finland) using photostimulable phosphor plates, which were scanned in a Digora PCT scanner (Soredex). For information about exposure settings and for more details, see Wiese et al.15 The skulls were placed on a photostat device to ensure correct positioning. The lateral tomograms were obtained with the mandible in a stable dental occlusion. Silicone impression was used to obtain a stable occlusion in case of tooth loss. The frontal tomograms were performed with the condyle positioned below the articular tubercle. A radiographer performed the tomographic examinations. Three pre-calibrated observers examined the lateral and the frontal images from both radiographic modalities individually. On lateral images, the condyle was assessed for the presence of attening, defect and osteophyte, whereas the mandibular fossa and the articular tubercle were assessed for the presence of attening and defect. On frontal images, only the condyle and the articular tubercle were assessed for the presence of attening and defect. Flattening was dened as loss of convexity of the bone outlines, defect as a local area of rarefaction in the layer of compact bone and osteophyte as a local outgrowth of bone arising from the mineralized surface.16 Multiple examples of attening, defects and osteophytes in TMJ tomograms were collected in an atlas, which the observers could consult in case they were in doubt how manifest changes should be scored as present. The CBCT images were assessed in the NewTomw 3G software on a 19 inch Philips at screen monitor, whereas the tomograms were assessed in the Digora software on a 19 inch IBM CRT monitor. Both monitors were placed in rooms with subdued lighting. The observers could use the enhancement facilities in the respective software as they liked. Two experienced oral radiologists (Observer1 and Observer 2) and one general dentist under specialist training (Observer 3) independently examined the lateral and the frontal images of each joint obtained with the two modalities. The left and the right joints in each skull were examined independently of each other, and lateral and frontal cross-sectional views of the same joint were also assessed independently of each other. A binary registration scale was used with the following scores: 0, not certain of change; 1, certain of change. 15 randomly selected joints were assessed a second time (3 4 weeks after the rst assessment) with both radiographic modalities to evaluate the intraobserver reproducibility. Several months prior to the radiographic examinations, the joints had been validated for the presence of
Dentomaxillofacial Radiology

194

TMJ radiography H Hintze et al

Table 1 Frequency of morphological changes in accordance with the gold standard Flattening % (n ) Condyle Mandibular fossa Articular tubercle 40.3 (64) 0.6 (1) 3.8 (6) Defects % (n ) 58.5 (93) 4.4 (7) 15.1 (24) Osteophytes % (n ) 8.2 (13) Not assessed Not assessed Total 159 159 159

morphological changes by the same three observers. A calibration session preceded the individual observer examination. The surfaces of the skulls condyles, mandibular fossae and articular tubercles were examined by naked-eye inspection for the presence of macroscopic attening, defect and osteophyte using a magnifying viewer with built-in light. Score 0 indicated not certain of change and Score 1 indicated certain of change. A change was dened as present if reported by at least two of the three observers. The naked-eye inspection served as the gold standard for the radiographic examinations. Data analysis The diagnostic accuracy for the detection of attening, defects and osteophytes in the various joint components on lateral and frontal views separately was expressed as the sensitivity and specicity, and differences between the two radiographic modalities were tested by paired t-test (SPSS package, GLM, version 10.0 for Windows; SPSS Inc., Chicago, IL). To express the general diagnostic accuracy for all changes in each joint component and in all joint components as a whole, the sum of non-matching gold standard and radiographic scores was used. The sum of non-matching scores was calculated for each possible change by counting the number of score differences of 1 and 2 1 (subtracting the radiographic score from the gold standard score). A score difference of 1 expressed a false-negative recording whereas a score difference of 2 1 expressed a false-positive recording. The sum of
250

differences was termed the absolute difference and was calculated for lateral and frontal views separately, and for the two projections in combination. To evaluate the accuracy of the combined views, attening and defects, which could appear in both projections, were dened to be present if they were recorded just from one of the projections, whereas changes appearing in one projection only were included in case they were recorded from the relevant projection. The lower the absolute difference, the higher the diagnostic accuracy. Differences between modalities were tested by paired t-test and differences between observers were tested by variance analysis (ANOVA). The level of signicance was P , 0.05. Intraobserver reproducibility for recordings obtained from lateral and frontal views in combination was expressed as kappa values.

Results One joint missed conventional tomography in the frontal plane and was excluded from the study sample, which thereafter included 159 joints. The frequency of morphological changes according to the gold standard in the condyles, mandibular fossae and articular tubercles is shown in Table 1. Since the number of changes in relation to the mandibular fossa was very low, the results for this joint component were excluded. In Figure 1, the mean absolute difference of all morphological changes in relation to the condyle and the articular tubercle is shown. On lateral views alone and on lateral and frontal views in combination, the absolute difference with CBCT was higher than with conventional spiral tomography, whereas the opposite was found using frontal views alone. However, only the difference found on lateral views alone was signicant (P 0.044). The diagnostic accuracy of changes in the condyle (Table 2) showed no signicant difference between the two modalities

200 Mean absolute difference

150

100

50

0 Lateral views Frontal views

CBCT Tomo

Lateral+Frontal views

Figure 1 Mean observer diagnostic accuracy for the detection of all morphological changes in the condyles and the articular tubercles on lateral views, frontal views and lateral and frontal views in combination using cone beam CT (CBCT) and conventional spiral tomography (Tomo) (*signicant difference, P 0.044)
Dentomaxillofacial Radiology

TMJ radiography H Hintze et al

195

Table 2 Diagnostic accuracy for the detection of morphological changes in the condyle using cone beam CT (CBCT) and conventional spiral tomography (Tomo) and P-values for their comparison Lateral projection Flattening CBCT Mean sensitivity Mean specicity 0.23 0.87 Tomo 0.11 0.95 P-value 0.21 0.22 Defect CBCT 0.15 0.95 Tomo 0.15 0.97 P-value 0.83 0.18 Osteophyte CBCT 0.15 0.96 Tomo 0.15 0.99 P-value 0.43 Frontal projection Flattening CBCT 0.40 0.90 Tomo 0.23 0.94 P-value 0.52 0.25 Defect CBCT 0.20 0.96 Tomo 0.21 0.95 P-value 0.94 0.67

on both lateral and frontal views. The mean observer sensitivities for the various types of changes were generally low, since they ranged from 0.11 for attening on lateral conventional spiral tomography views to 0.40 for defects on frontal CBCT views. In contrast, the mean specicities for the same changes were high, ranging from 0.87 for attening on lateral CBCT views to 0.99 for osteophytes on lateral conventional spiral tomography views. Mean sensitivity and specicity values for changes in relation to the articular tubercle are shown in Table 3. Also in this joint component, no signicant differences in diagnostic accuracies were found between CBCT and conventional spiral tomography. Only for the detection of defects on frontal views did conventional spiral tomography show a signicantly higher specicity than CBCT. The tendency for low senstivity and high specicity values as found for changes in the condyle was valid also for the articular tubercle. In particular, defects were very difcult to identify. The individual observer performance with the two radiographic modalities using lateral and frontal views in combination can be seen in Figure 2. No signicant differences in observer performances were found for the detection of changes in relation to the condyle and the articular tubercle, or the two components assessed together with any of the two radiographic modalities. The most manifest observer variation was seen for assessment of condylar changes with conventional spiral tomography and here the variation was mainly caused by large (non-signicant) variations for the detection of bone defects. Kappa values for intraobserver reproducibility for the detection of morphological changes in relation to the condyle and the articular tubercle assessed together from lateral and frontal views in combination ranged as follows for CBCT and conventional spiral tomography: Observer 1: 0.55 1.00 and 0.66 1.00; Observer 2: 0.32 1.00 and 0.32 1.00; Observer 3: 1.00 (no range) and 0.87 1.00.

Discussion In general, tomography is considered the most accurate of the traditional radiographic techniques for imaging of the TMJ.17 20 To obtain the most realistic images of the joint components, the examination ought to be orientated perpendicular and/or parallel to the long axis of the condyle. This requires a pre-examination of the joint from which the angulation of the long axis can be assessed. This, in addition to the fact that multiple tomographic cuts are often needed to examine the joint from the most lateral to the most medial part in the lateral projection and from the most anterior to the most posterior part in the frontal projection, might well result in high patient radiation doses. Moreover, the examination procedure for a bilateral TMJ tomography in two planes is time-consuming and occupies the radiographic unit for a long period of time.14 In case the diagnostic accuracy of TMJ changes is not jeopardized, many advantages could be obtained with the use of the CBCT technique. A CBCT examination with the NewTomw 3G scanner is denitely much shorter, includes image data for both the right and left TMJs from a single 3608 rotation scan around the patients head and simplies patient positioning, and the radiation dose may be lower than for multiple-cuts tomography. The latter has not been evaluated yet, but from a recent study the radiation dose needed for a 9 inch NewTomw 9000 scan was found to correspond to the dose from two to four or from three to seven panoramic radiographs, depending on whether the effective organ doses included or excluded the salivary glands.21 In comparison with a panoramic exposure, a onecut tomography examination performed in the posterior part of the jaws using rectangular collimation has been found to require about the same dose.22,23 In the present study, an average of four tomographic cuts in both the lateral and frontal planes were exposed for each TMJ. In addition, scout images preceding the actual tomography were needed and obtained. All in all, a conventional bilateral TMJ

Table 3 Diagnostic accuracy for the detection of morphological changes in the articular tubercle using cone beam CT (CBCT) and conventional spiral tomography (Tomo) and P-values for their comparison (signicant value is marked in bold) Lateral projection Flattening CBCT Mean sensitivity Mean specicity 0.28 1.00 Tomo 0.39 0.99 P-value 0.52 0.74 Defect CBCT 0.03 0.98 Tomo 0.03 0.98 P-value 1.00 0.42 Frontal projection Flattening CBCT 0.12 1.00 Tomo 0.17 1.00 P-value 0.87 0.42 Defect CBCT 0.07 0.96 Tomo 0.06 0.99 P-value 0.42 0.02

Dentomaxillofacial Radiology

196
200

TMJ radiography H Hintze et al

Absolute difference

150

100

50

0
CBCT-Condyle Tomo-Condyle CBCT-Tubercle Tomo-Tubercle CBCTTomoCondyle+Tubercle Condyle+Tubercle

Observer 1

Observer 2

Observer 3

Figure 2 Diagnostic accuracy of each observer for the detection of morphological changes in the condyles, the articular tubercles and the condyles and the articular tubercles assessed together on lateral and frontal views in combination using cone beam CT (CBCT) and conventional spiral tomography (Tomo)

tomographic examination might well require a much higher dose than needed for CBCT. From the present results, no general signicant difference in diagnostic accuracy was found between CBCT and conventional spiral tomography. Evaluation of the two joint components separately showed the same result, with the exception of the detection of bone defects in the articular tubercle on frontal views, where the specicity was signicantly higher with the conventional spiral tomography modality than with CBCT. The difference in specicity was, however, only 3% and such a small variation will probably not have any consequences in the clinic. For both radiographic modalities, it was striking to see how ineffective they were for the identication of attening, defects and osteophytes. Condylar attening was present in 40% of the 159 examined joints in accordance with the gold standard but, from lateral views, on average only 23% was identied correctly with CBCT and on average 11% with conventional spiral tomography. Using frontal views for assessment of the same change, this percentage increased to on average 40% with CBCT and on average 23% with conventional spiral tomography, but still the percentages were disappointingly low. However, the higher sensitivity obtained from frontal views ought to motivate the clinician to order such a projection as a supplement to the lateral view when the TMJ is to be examined. This recommendation is in agreement with conclusions from previous studies on effective projections for the assessment of the TMJ.14,17,24,25 However, frontal views cannot replace lateral views since condylar anterior osteophytes might be visible on lateral views only. The insignicant differences between sensitivity values for the detection of condylar attening with the two modalities were not ascribed to the different positions of the skulls mandibles during examination. If
References
1. Preda L, Di Maggio EM, Dore R, La Fianza A, Solcia M, Schino MR, et al. Use of spiral computed tomography for multiplanar dental reconstruction. Dentomaxillofac Radiol 1997; 26: 327 331.
Dentomaxillofacial Radiology

this had had an inuence on the recordings of the images, variations between the accuracy values for the other morphological changes would also be expected, but such differences were not found. For the identication of condylar defects and osteophytes, the sensitivities were lower than the values for attening, whereas the specicities were high. This result indicates that only a minority of truly present condylar defects and osteophytes are correctly identied, while the majority of normal condyles (without defects and osteophytes) are correctly identied as sound. This nding was also valid for attening and defects in the articular tubercle. For a general assessment of the TMJ, no signicant differences in diagnostic accuracy were found between the observers with either of the two modalities. The variation in intraobserver agreement for all three observers for the assessment of the various changes in the two joint components was about the same with both modalities. This observation was somewhat surprising since the observers were rather familiar with conventional tomograms, but completely unfamiliar with CBCT for TMJ examinations. This might indicate that NewTom images are very reader-friendly and easy to become familiar with. In conclusion, generally no signicant differences in diagnostic accuracy for the detection of morphological bone changes in relation to the condyle and the articular tubercle were found between images from cone beam CT obtained with the NewTomw 3G scanner and conventional cross-sectional tomography obtained with the Cranex Tome unit. If the present results are also valid for patients, the modality requiring least radiation dose and being most convenient for the patient and the radiographic staff ought to be used.

2. Cohnen M, Kemper J, Mobes O, Pawelzik J, Modder U. Radiation dose in dental radiology. Eur Radiol 2002; 12: 634 637. 3. Danforth RA, Dus I, Mah J. 3-D volume imaging for dentistry: a new dimension. J Calif Dent Assoc 2003; 31: 817 823.

TMJ radiography H Hintze et al

197

4. 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; 95: 371377. 5. Schulze D, Heiland M, Thurmann H, Adam G. Radiation exposure during midfacial imaging using 4- and 16-slice computed tomography, cone beam computed tomography systems and conventional radiography. Dentomaxillofac Radiol 2004; 33: 83 86. 6. Pawelzik J, Cohnen M, Willers R, Becker J. A comparison of conventional panoramic radiographs with volumetric computed tomography images in the preoperative assessment of impacted mandibular third molars. J Oral Maxillofac Surg 2002; 60: 979 984. 7. Mengel R, Candir M, Shiratori K, Flores-de-Jacoby L. Digital volume tomography in the diagnosis of periodontal defects: an in vitro study on native pig and human mandibles. J Periodontol 2005; 76: 665 673. 8. Hatcher DC, Dial C, Mayorga C. Cone beam CT for pre-surgical assessment of implant sites. J Calif Dent Assoc 2003; 31: 825 833. 9. Lascala CA, Panella J, Marques MM. Analysis of the accuracy of linear measurements obtained by cone beam computed tomography (CBCT-NewTom). Dentomaxillofac Radiol 2004; 33: 291 294. 10. Aranyarachkul P, Caruso J, Gantes B, Schulz E, Riggs M, Dus I, et al. Bone density assessments of dental implant sites: 2. Quantitative cone-beam computerized tomography. Int J Oral Maxillofac Implants 2005; 20: 416 424. 11. Heiland M, Schulze D, Rother U, Schmelzle R. Postoperative imaging of zygomaticomaxillary complex fractures using digital volume tomography. J Oral Maxillofac Surg 2004; 62: 1387 1391. 12. Walker L, Enciso R, Mah J. Three-dimensional localization of maxillary canines with cone-beam computed tomography. Am J Orthod Dentofacial Orthop 2005; 128: 418 423. 13. Honda K, Arai Y, Kashima M, Takano Y, Sawada K, Ejima K, et al. Evaluation of the usefulness of the limited cone-beam CT (3DX) in the assessment of the thickness of the roof of the glenoid fossa of the temporomandibular joint. Dentomaxillofac Radiol 2004; 33: 391 395. 14. Tsiklakis K, Syriopoulos K, Stamatakis HC. Radiographic examination of the temporomandibular joint using cone beam computed tomography. Dentomaxillofac Radiol 2004; 33: 196 201. 15. Wiese, M., Hintze, H., Svensson, P., and Wenzel, A. Comparison of diagnostic accuracy of lm and digital tomograms for assessment of morphological changes in the TMJ. Dentomaxillofac Radiol 2007; 36: 12 17.

, Petersson A. Radiography of the temporomandibular 16. Omnell K-A joint utilizing oblique lateral transcranial projections. Comparison of information obtained with standardized technique and individualized technique. Odont Revy 1976; 26: 77 92. 17. Tyndall DA, Davies KL, Ludlow JB. A comparison of zonography, multidirectional tomography, and biplanar panoramic images for detection of simulated condylar arthropathy of the temporomandibular joint. Int J Adult Orthod Orthognath Surg 1995; 10: 277 284. 18. Ludlow JB, Davies KL, Tyndall DA. Temporomandibular joint imaging. A comparative study of diagnostic accuracy for the detection of bone change with biplanar multidirectional tomography and panoramic images. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1995; 80: 735 743. kerman S. 19. Cholitgul W, Petersson A, Rohlin M, Tanimoto K, A Diagnostic outcome and observer performance in sagittal tomography of the temporomandibular joint. Dentomaxillofac Radiol 1990; 19: 1 6. 20. Brooks SL, Brand JW, Gibbs SJ, Hollender L, Lurie AG, Omnell K , et al. Imaging of the temporomandibular joint. A position paper of A the American Academy of Oral and Maxillofacial Radiology. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997; 83: 609 618. 21. Ludlow JB, Davies-Ludlow LE, Brooks SL. Dosimetry of two extraoral direct digital imaging devices: NewTom cone beam CT and Orthophos Plus DS panoramic unit. Dentomaxillofac Radiol 2003; 32: 229 234. 22. Frederiksen NL, Benson BW, Sokolowski TW. Effective dose and risk assessment from lm tomography used for dental implant diagnostics. Dentomaxillofac Radiol 1994; 23: 123 127. 23. Dula K, Mini R, van der Stelt PF, Sanderink GCH, Schneeberger P, Buser D. Comparative dose measurements by spiral tomography for preimplant diagnosis: the Scanora machine versus the Cranex Tome radiography unit. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001; 91: 735 742. 24. Sato H, Fujii T, Yamada N, Kitamori H. The contribution of frontal tomography to the diagnosis of temporomandibular joint osteoarthritis. Dentomaxillofac Radiol 1992; 21: 77 80. 25. Bouwman JPB, Kerstens HCJ, Tuinzing DB. Condylar resorption in orthognathic surgery. Oral Surg Oral Med Oral Pathol 1994; 78: 138 141.

Dentomaxillofacial Radiology