Understanding the TMJ morphology and its relative position in the class I, class II, and class III groups remains a challenge for clinicians. Knowledge on the spatial variations of normal condyle-glenoid fossa relationship could allow the clinician to potentially identify the beginning of a degenerative joint disease or indicate problems already established, as well as better treatment planning where obtaining values closer to normal is indicated [14, 15]. Therefore, the accurate determination of these values in conjunction with clinical observations could be of great importance for diagnosis and treatment planning in different skeletal relationships.
Ricketts  used laminographic images to analyze TMJs whose images had minimum distortion. Based on this, we used CBCT images (theoretically zero distortion)  in order to determine the condyle spatial position in relation to the articular eminence for the different skeletal anteroposterior relationships and to compare these values with those taken from laminographic images by Ricketts . In this study, we found that the condyles in the class II and class III groups are more superior than those in the class I group. Ricketts  and Katsavrias  mentioned that the condyles in class III skeletal relationship patients were closer to the roof of the glenoid fossa, which coincides with our results. Nevertheless, with regard to the class II group, Ricketts  stated that the condyle was positioned lower, which does not agree with our results, which may suggest that the condyle position is influenced by a variety of factors related to the included vertical pattern. The class II and class III groups selected for our study had a long vertical facial pattern. In this regard, Paredes et al.  mentioned that the upper distance between the condyle and fossa is increased in short facial types and reduced in long facial types, coinciding with our results. Therefore, both vertical patterns (long or short) should also be considered in determining condylar position clinically and radiographically. The clinical significance of the upper distance is that it determines whether the condyle is morphologically altered since an increase or decrease of the distance may produce pathologies such as resorption and condylar hyperplasia [14, 15].
Regarding the posterior distance, no significant differences were observed between the three groups studied. The posterior condylar position relative to a line perpendicular to the Frankfurt plane passing through the midpoint of the sagittal diameter of the external auditory canal may actually be influenced by the morphology and location of it . Differences were observed indicating that the condyles of the class II and III groups were anteriorly located when compared to the class I group. Similarly, Ricketts  in his analysis of variations in condyle position mentions that in class II malocclusions the condyle appears to be more anterior. We suggest that future studies take into account the posterior wall of the glenoid fossa instead of the line perpendicular to the Frankfurt plane passing through the midpoint of the sagittal diameter of the external auditory canal. In our study, the parameters were determined by the criteria used by Ricketts.
With regard to the anterior distance from the condyle to the eminence, there is a statistically significant difference between the class I and the class II groups, and between the class I and class III groups, suggesting that the condyle in the class II group is positioned more anteriorly, coinciding with the results of Katsavrias , who stated that in class II groups of both divisions, the condyle is located in a more anterior position. Pullinger and Hollender  showed that a non-concentric position of the condyle is a characteristic of class II malocclusion and that the condyles are positioned further in the anterior direction in patients with class II malocclusion than in class I patients. Furthermore, our study showed that there is a statistically significant difference in the anterior distance between the class I and III groups, finding that the condyle in the class III group is in a more anterior position compared with the class I group. This result may be due to a tendency towards a non-concentric position of the condyle-glenoid fossa, morphological variability, and the degree of severity of skeletal and vertical pattern type. Rodrigues et al.  evaluated the concentric position in the mandibular fossa and found a non-concentric position on both right and left sides for class II and class III groups. Katsavrias  showed that variations in the morphology of the condyle are related principally to the inclination of the head of the condyle and that the shapes of the condyle and fossa are different in the class II and class III groups. It is important to note that if this distance decreases with respect to the normal distance and the treatment provided further decreases, this distance could face condylar distraction, highlighting the clinical importance of knowing the TMJ without alteration and according to the skeletal pattern.
With respect to the articular eminence angle, Katsavrias  and Sülün et al.  showed that a pronounced inclination in the articular eminence can predispose a dysfunctional temporomandibular joint and that class II division 2 malocclusion cases are characterized by strong and high articular eminences. However, in this investigation, the articular angle was found to be highest in the class I group (58°), followed by class II (51°), and lowest in the class III group (42°). We cannot, therefore, confirm that a pronounced inclination in the articular eminence can influence the development of the TMJ. Christiansen et al.  reported that the normal value of the articular eminence angle in adults is 30° to 60°, while the values measured in this study, for young adult patients, varied between 42° and 58°. Ricketts  says that at the age of 7.5 years the inclination is 46°, and at 12.5, 18.5, and 22 years of age, the inclination is 52°, 57°, and 59°, respectively. In this regard, we would point out that the sample used in this study had a mean age of 25 years and that the mean inclination of the eminence of the study groups was 50° with a standard deviation of ±8°, similar to the values defined by Ricketts . The ratio of the inclination of the articular eminence within different skeletal pattern types can be affected by other factors related to age, sex, dental occlusion, and incisor or canine guidance, among other factors. Most of these variables were not considered in our study.
The articular eminence height was found to be lower in the class III group when compared with the class I group. This result could be due to the influence of facial vertical pattern length and severity of the skeletal relationship. Regarding this, Katsavrias  mentions that variations in the shape of the fossa are related to the inclination and height of the eminence and stated that in the class III group the condyle is higher or closer to the roof of the fossa. Vitral et al.  did not find any significant differences in the height of the glenoid fossa between the class I and class II groups, results that coincide with those obtained in this study. Moreover, Ricketts , Cohlmia et al. , and Vitral et al.  found no correlation between the depth of the glenoid fossa, the inclination of the eminence, and the different dental malocclusions.
The values of the spatial analysis of condylar position in CBCT images of the three study groups, class I, class II, and class III (Figure 9), are within the ranges shown by Ricketts in his study of condylar position in laminographic images. These similarities maybe due to the low image distortion in laminography and the relative absence of distortion in CBCTs. It has to be noted that the measurements obtained in this study should be more accurate because the CBCT images are isotropic.
The attempt of this initial study was to evaluate the precision of TMJ internal linear and angular measurements as measured with CBCT images. An important limitation regarding temporomandibular dysfunction (TMD) diagnosis was identified. In this study, it was not possible to measure the daily and chronic stress of patients, habits, pathological and functional changes, and muscle activity that have been shown to impact TMD.
Although CBCT maybe one useful tool for measuring TMJ bony structures given the improved anatomical resolution it provides, one should not forget the ALARA principles when determining the need for further images and related increased ionizing radiation. Careful consideration of a cost/benefit analysis should be encouraged. Also, it should be noted that although the results obtained in this study show statistically significant differences, the clinical relevance of the differences is questionable.