Comparative Evaluation of Temporomandibular Condylar Changes Using Texture Analysis of CT and MRI Images

Abstract

This study aims to compare computed tomography (CT) with magnetic resonance imaging (MRI) of the temporomandibular joint (TMJ) by using texture analysis (TA) to detect condylar bone marrow changes associated with the flattening and erosion of cortical bone. A total of 47 patients from the Dentomaxillofacial Radiology Division at São Paulo State University were evaluated. Images from 250 CT and 250 MRI images were assessed by experienced radiologists employing OnDemand3D software. Texture parameters were extracted with MaZda software (version 4.6), and we focused on regions of interest within the condyles. Statistical analysis revealed significant differences in texture parameters between the affected and control groups. CT images showed higher correlation values in cases of flattening, whereas MRI images demonstrated substantial changes in texture parameters for both flattening and erosion. These findings suggest that the texture analysis of CT and MRI images can effectively detect early and advanced degenerative changes in the TMJ, thus providing valuable insights into the underlying pathophysiology and aiding in early intervention and treatment planning.

1. Introduction

The morphology of the condyle of the temporomandibular joint (TMJ) changes according to age, gender, facial type, functional load, occlusal force, and type of malocclusion, and between the left and right sides [1]. Parafunctional factors such as bruxism and clenching overload the joints, leading to remodeling [1,2] and articular disc derangement [3].

The trabecular bone organization is characterized by porosity, trabecular thickness, and anisotropy, which can be altered depending on the movements resulting from chewing [4]. The remodeling of the condylar cartilage of the mandible occurs in response to mechanical deformations, in which chondrogenesis and endochondral ossification are regulated to obtain a better balance between mechanical stress and the load capacity of the joint [5]. Among the degenerative signs affecting the condyle, one can cite flattening, formation of osteophytes, presence of subchondral sclerosis, and erosion of cortical bone [6,7], with flattening being a condition that reflects an initial process and cortical erosion occurring in a more advanced stage. The destruction of the condyle can cause malformations such as retrognathism, anterior open bite, and facial asymmetry [7]. Addressing this condition in its early stages is crucial for several reasons:

• Prevention of progressive deformity: Early intervention can halt or slow the progression of facial asymmetry and malocclusion, potentially avoiding more complex surgical corrections in the future [8];

• Preservation of growth potential: In younger patients, timely treatment may allow for more normal facial growth and development [9];

• Functional improvement: Early management can help maintain or restore proper jaw function, including mastication and speech [10];

• Psychosocial benefits: Preventing severe facial deformities can have significant positive impacts on a patient’s self-esteem and social interactions, particularly in adolescents [11];

• Simplified treatment: Early detection and treatment may allow for less invasive procedures and potentially better outcomes compared to addressing advanced deformities [12].

Computed tomography (CT) has been used to evaluate TMJ since the early 1980s as it offers superior image quality for assessing bone structures. Therefore, CT is a highly recommended examination in the evaluation of condyles, particularly in the diagnosis of ankyloses, arthritis, and osteoarthritis [13].

Magnetic resonance imaging (MRI) is considered the most accurate method for showing structural alterations, particularly in the TMJ soft tissues, as it can identify the articular disc [9]. MRI is essential to detect changes in the condylar bone marrow, such as avascular necrosis and edema [14].

Radiomics is a field of medical imaging dedicated to quantitative analysis of original medical images [15]. Features extracted from the images (e.g., shape, texture, and pixel intensity) and their analysis are used to obtain detailed information about the morphology and heterogeneity of a lesion or tissue [15,16]. When combined with other clinical data, this information can be valuable for the diagnosis, treatment, and prognosis of a disease [16].

Although MRI is the gold standard for detecting bone marrow changes, identifying these changes through a visual inspection of images remains challenging due to the limitations of the human eye in distinguishing subtle differences in grayscale intensities [17]. Therefore, incorporating radiomics into the methodology enhances the MRI contrast, which allows for a more detailed and quantitative analysis of imaging features. This significantly improves the detection and evaluation of tissue and lesion heterogeneity [15].

Texture analysis (TA) is a statistical image analysis technique based on the use of radiomics, in which the distribution of pixel signals is quantitatively assessed by comparing them to neighboring pixels in a non-invasive way. TA has been developed by measuring the distribution of grayscale levels in the region of interest (ROI) delineated in the image, thus distinguishing lesions from healthy tissues [18,19,20,21,22]. Therefore, data from TA in combination with clinical information can indicate and validate diagnostic hypotheses [23].

One of the most frequent methods for extracting texture parameters from grayscale images is the co-occurrence matrix (GLCM) [24], which can detect subtle changes in images [25,26,27]. In recent years, computerized analyses of morphology and texture have been used to aid in the diagnosis of various pathologies [21,22,23,24]. Automated patterns applied to medical image analysis for lesion recognition have high accuracy and are recognized as imaging biomarkers [28].

The objective of this study was to compare CT and MRI images of TMJs affected by the flattening and erosion of the cortical bone using TA to detect condylar bone marrow changes in these joints. By applying TA to compare CT and MRI images of TMJs affected by flattening and erosion of the cortical bone, we aim to enhance the detection of subtle condylar bone marrow changes that are often challenging to identify through visual inspection alone. This approach has the potential to improve early diagnosis of TMJ degeneration, enabling more timely interventions. This research also seeks to establish quantitative imaging biomarkers specific to TMJ degeneration, which could help standardize assessment criteria across the field. By focusing on these specific conditions, this study may provide new insights into the progression of TMJ degeneration, aiding clinicians in distinguishing between various stages of the condition. Furthermore, the findings from this study could inform the development of more targeted treatment strategies and contribute to the advancement of computer-aided diagnostic tools in TMJ imaging, potentially enhancing the accuracy, efficiency, and early detection of condyle changes in clinical practice.

2. Materials and Methods

This is a retrospective study. All procedures in this study were conducted in full accordance with the ethical principles set by the Helsinki Declaration of 1975, as revised in 2013. All patients had provided written consent after being informed about the use of CT and MRI images.

This study was approved by the institutional review board of the School of Dentistry of the University of São Paulo (USP) according to protocol number 56631222.9.0000.0075.

All the patients enrolled in the study had been admitted for CT and MRI for the evaluation of TMJ complaints. The database held by the Dentomaxillofacial Radiology Division of the São José dos Campos School of Dentistry, UNESP, was reviewed between January 2019 and April 2021.

2.1. Image Acquisition

Non-contrast enhanced CT examinations were performed by using a 4-channel multi-detector CT scanner (Alexion, Toshiba/Canon, Otawara, Japan) with contiguous 1 mm thick slices at 1 mm intervals, operated with 100 kV, 100 mA, 1 s/rotation, 512 × 512 matrix, a gap measuring 0.8 mm, a voxel size of 0.37 mm × 0.37 mm, and a field of view (FOV) of 180 mm × 180 mm.

MRI scans were acquired on a 1.5 Tesla scanner (Sigma, General Electric, Milwaukee, WI, USA) with a dedicated TMJ surface coil measuring 0.6 m in diameter. Sagittal T1 images were obtained in the closed-mouth position (TE 8.5 ms and TR 850 ms), with an FOV of 150 mm × 150 mm, 2 mm thickness, a 1.0 mm intersection gap, and a raw data matrix measuring 512 × 512 mm.

The digital imaging and communications in medicine (DICOM) format was used to export all image data acquired from CT and MRI scanners.

2.2. Image Analysis

Patients with both CT and MRI images acquired within a month were included after being retrieved from the computer database for assessment. A total of 500 examinations, consisting of 250 CT images and 250 MRI images, were analyzed by using OnDemand3D software (version 1.0.9.3223, CyberMed Inc., Seul, Republic of Korea). Two dentomaxillofacial radiologists, with more than 15 years of experience, selected and analyzed the images based on a consensus on bone changes, namely erosion and flattening.

Initially, they considered including cases among those in the sample related to other alterations, such as the presence of osteophytes and subchondral cysts. However, the decision was made to focus exclusively on cases with flattening and erosion of the condylar cortex. This selection was based on the rationale that flattening represents an initial stage of degenerative changes, whereas erosion indicates a more advanced and severe stage. Additionally, the number of cases with other alterations in our sample was small, which could potentially have introduced a bias in the results and reduced the statistical power of our analysis. By concentrating on these two specific alterations, we aimed to capture the progression of TMJ degeneration more clearly to provide a comprehensive understanding of the early and late manifestations of the condition. This approach allowed for a more precise evaluation of the efficacy of TA in identifying varying degrees of degenerative changes, thus enhancing the clinical relevance of our findings. Therefore, the final sample consisted of 47 patients of both genders, which was divided into groups of TMJs with flattening and erosion. TMJs without any alteration constituted the control group.

The inclusion criteria were patients from both genders older than 16 years who presented with bilaterally visible TMJs, whereas exclusion criteria were patients with images showing severe artifacts or a history of systemic diseases, maxillofacial trauma, recent jaw surgery, or congenital bone/cartilage disease.

After applying the inclusion and exclusion criteria, 47 patients aged between 16 and 72 years old were included in the study.

2.3. Extraction and Analysis of Texture Features

TA analysis was performed by a different fellowship-trained dentomaxillofacial radiologist with five years of CT and MRI experience who was also unaware of the final diagnosis. The radiologist manually selected the first sagittal slice in the most central region in all cases, where the condyle was more visible. Two additional sequential parallel sections, one more medial and another more distal, were also analyzed. Each DICOM slice was processed and converted into bitmap format by using OnDemand3D software (CyberMed Inc., Seoul, Republic of Korea). The same radiologist chose the region of interest (ROI) by drawing lines that intersected the outer and inner edges of the condyles in the sagittal view as well as the medial and lateral borders in the sagittal perspective. Next, to enhance precision in locating the center of the condyles, a line was drawn in the middle of these aforementioned lines to guide the outline of the ROI with a diameter of 6.5 mm at the center of the condyle (Figure 1). TA analysis of CT and MRI images was based on parameters (i.e., features) extracted by using MaZda software, version 4.6 (Technical University of Lodz, Poland), on a 15-inch MacBook Pro notebook (Apple, Los Altos, CA, USA) with an Intel^® CoreTMi5 (Intel, Santa Clara, CA, USA), 2.4 GHz, 4 GB RAM, 1067 MHz, DDR3 processor (Samsung, Suwon, Republic of Korea) and Microsoft Windows version 10. The same diameter of the ROI was used in the three sections to create volumetric data, that is, values of the volume of interest (VOI) for texture analysis [29]. The texture parameters from each of the three sections were extracted, and the mean value was calculated for each condyle.

GLCM is a statistical method for the calculation of the properties of spatial relationships between pixels [29]. Texture parameters were computed from the GLCM, corresponding to distances of 1, 2, and 3 pixels in the four directions of the image (i.e., horizontal, diagonal, vertical, and anti-diagonal, corresponding to 0°, 45°, 90°, and 135°, respectively).

To minimize the factors at play and select seven parameters with a stronger connection with bone tissue, we used a recent study as a basis to select seven texture parameters [29]. The Supplementary Table S1 shows the texture features computed by using MaZda software.

To provide a comprehensive overview of our methodological approach, Figure 2 presents a visual representation of the entire pipeline, from image acquisition to texture analysis. This flowchart illustrates the key steps in our process, including image acquisition, analysis and selection, preparation for texture analysis, feature extraction, and data analysis.

2.4. Statistical Analysis

Statistical analyses were performed by using R software, version 3.6.0 (R Foundation for Statistical Computing, Vienna, Austria). Spearman’s correlation coefficient was used to assess the correlation between distances of the same texture parameter. Groups with and without disease were compared by using the Mann–Whitney test. All statistical analyses were performed at a significance level of 5%.

3. Results

This study included 47 patients of both genders (74% females and 26% males), aged between 16 and 72 years old (mean age of 32.5 years). Texture parameters were measured for each patient by using both CT and MRI images of the condylar bone marrow. Table 1 shows the distribution of the alterations (i.e., flattening and erosion) in the sample. Concerning the pathologies of flattening and erosion, the analysis of the right and left sides was performed as the same patients did not necessarily have the condition on both sides.

Texture parameters were obtained for each patient by using CT and MRI images of both the right and left sides.

Seven parameters (angular second moment [AngScMom], contrast, correlation [Correlat], entropy, inverse difference moment [InvDfMom], sum of entropy [SumEntrp], and sum of squares [SumOfSqs]) were extracted for 12 positions. Before calculating the average of the 12 positions of each parameter, Spearman’s correlation coefficient between them was calculated. As a high correlation was observed between the positions in six of the seven parameters, we decided to use the average between them. Because no strong association was observed between the positions for the correlation parameter, all 12 parameters were analyzed separately. Figure 2 shows the correlation matrix calculated for CT and MRI images. The two distances were organized according to four directions in the following positions: S10 (d1 = 1; ∡ = 0°), S01 (d1 = 1; ∡ = 45°), S11 (d1 = 1; ∡ = 90°), S1m1 (d1 = 1; ∡ = 135°), S20 (d2 = 2; ∡ = 0°), S02 (d2 = 2; ∡ = 45°), S22 (d2 = 2; ∡ = 90°), S2m2 (d2 = 2; ∡ = 135°), S30 (d3 = 3; ∡ = 0°), S03 (d3 = 3; ∡ = 45°), S33 (d3 = 3; ∡ = 90°), S3m3 (d3 = 3; ∡ = 135°).

Figure 3 and Figure 4 show a comparison of the groups with and without pathology regarding the texture parameters of CT and MRI images for both sides. The groups were statistically different at a 5% significance level for the parameters marked with an asterisk (*). The parameters contrast and SumOfSqs were divided by 10 to be included in the graphs without distorting the scale for the other parameters.

Statistically significant differences in texture parameters between the groups with and without TMJ pathologies (flattening and erosion) for CT and MRI images are shown in Table 2.

Texture parameters included angular second moment [AngScMom], contrast, correlation [Correlat], entropy, inverse difference moment [InvDfMom], sum of entropy [SumEntrp], sum of squares [SumOfSqs].

4. Discussion

This study yielded compelling results by shedding light on the potential of TA in detecting early signs of TMJ degeneration. Texture parameters were meticulously measured for each patient by using both CT and MRI modalities with a focus on the condylar medulla.

In the cases of condylar flattening, which is a precursor to more advanced degenerative changes, TA revealed significant differences between affected and unaffected groups. Interestingly, the group with flattening showed higher correlation values, particularly in the positions S(1,0), S(1,−1), S(2,0), S(2,−2) and S(3,0), as observed in the CT images. These findings were consistent for both the right and left condyles, thus underscoring the sensitivity of TA in detecting early degenerative changes.

MRI-based TA also yielded notable results, as the group with flattening on the left condyle showed higher correlation values in S(3,3). As the degenerative process progresses, condylar erosion becomes a prominent feature. In this study, TA revealed distinct patterns associated with erosive changes. CT images of the right condyle in the group with erosion showed higher contrast values, a higher sum of squares, and lower correlation values in S(2,2).

On the left condyle, the group with erosion showed higher correlation values in the positions S(0,1), S(0,2), and S(0,3), as detected via CT images.

MRI-based TA also contributed with valuable insights, in which the group with erosion on the left condyle showed a lower angular second moment, higher contrast, higher sum of squares, lower inverse difference moment, higher sum of entropy, and higher entropy values.

These findings highlight the sensitivity of TA in detecting both early and advanced degenerative changes in TMJ condyles based on complementary information from CT and MRI images.

The textural patterns observed in this study provide valuable insights into the pathophysiological processes underlying TMJ degeneration. Higher correlation values associated with condylar flattening and erosion suggest higher linear interdependence between pixel intensities, which reflects changes in bone density and organization.

As the degenerative process progresses, the body attempts to adapt by altering the trabecular architecture and increasing bone density in the affected regions [30]. This adaptive response manifests as higher correlation values in TA, indicating a more organized and interdependent distribution of pixel intensities.

Furthermore, the observed changes in the contrast, sum of squares, and entropy parameters in the cases of condylar erosion reflect alterations in the heterogeneity and complexity of the medullary bone structure [31,32]. These findings align with the expected pathophysiological changes associated with advanced degenerative processes, such as bone remodeling, trabecular disruption, and the formation of erosive lesions [33].

This study’s findings underscore the complementary nature of CT and MRI in assessing TMJ degeneration through TA. While CT excels in capturing changes related to cortical bone integrity and density, MRI provides valuable insights into alterations within the medullary bone and surrounding soft tissues.

By combining the strengths of both imaging modalities, TA offers a comprehensive evaluation of the degenerative process by encompassing cortical and medullary bone changes, including associated soft tissue alterations. This multimodal approach enhances diagnostic accuracy and facilitates a more holistic understanding of the disease’s progression.

The ability to detect early degenerative changes within TMJ condyles through TA holds significant clinical implications. Early intervention is essential in mitigating the progression of TMJ degeneration and preventing irreversible damage to the joint [34,35].

By identifying subtle alterations in the medullary bone structure, TA can serve as a valuable tool for clinicians in enabling them to initiate appropriate treatment strategies at the earliest possible stage. This proactive approach may include conservative measures, such as lifestyle modifications, physical therapy, or pharmacological interventions, aimed at slowing or halting the degenerative process.

Moreover, the quantitative nature of TA parameters opens the door for personalized treatment planning and monitoring. By tracking changes in these parameters over time, clinicians can tailor treatment regimens to individual patient needs, thus optimizing outcomes and minimizing potential complications.

As the field of medical imaging continues to evolve, the integration of artificial intelligence (AI) and machine learning (ML) techniques holds immense potential for further enhancing the diagnostic capabilities of TA [36].

By leveraging large datasets of TA parameters and clinical outcomes, AI algorithms can be developed to recognize intricate patterns and correlations that may be imperceptible to human observers [37]. These algorithms can then assist in the automated detection, classification, and prognosis of TMJ degenerative conditions by streamlining the diagnostic process and reducing the potential for human error.

Additionally, the incorporation of ML techniques can facilitate the development of predictive models to enable clinicians to anticipate the progression of degenerative changes and tailor treatment strategies accordingly. This proactive approach can significantly improve patient outcomes and quality of life.

Our study’s methodology aligns with and builds upon recent advancements in TA for TMJ evaluation. Girondi et al. [38] applied TA to MRI to identify changes in TMJ discs affected by effusion, utilizing similar GLCM parameters in MaZda software. Their findings of distinct texture patterns in effusion-affected discs support the potential of TA in detecting subtle TMJ changes. In a related study, Ricardo et al. [39] employed MRI-based TA for the quantitative evaluation of the mandibular condyle in juvenile idiopathic arthritis, demonstrating the technique’s utility in assessing bone changes in inflammatory conditions. Furthermore, Nussi et al. [40] extended the application of TA to cone beam computed tomography (CBCT) images of the mandibular condyle, correlating texture features with gender and age. These studies collectively underscore the growing importance of TA in TMJ research and its potential to enhance diagnostic accuracy across various imaging modalities. Our work contributes to this emerging field by focusing specifically on TMJ degeneration, thereby expanding the application of TA in TMJ disorders and reinforcing its value as a quantitative tool in radiological assessment.

Despite the promising results, this study has several limitations that need to be addressed. Firstly, the relatively small sample size limits the robustness and generalizability of our findings; a larger cohort would provide more comprehensive data. Secondly, the retrospective study design may introduce selection bias. Thirdly, while experienced radiologists performed image analysis, there is a potential annotation bias due to the inherent subjectivity in interpreting imaging findings, which could have impacted our results. This limitation underscores the need for future studies to implement measures for quantifying inter-rater reliability. Additionally, the study has primarily focused on imaging findings, and clinical correlation with symptoms, functional outcomes, and patient-reported measures is crucial to validate the clinical use of TA in the evaluation of TMJ degeneration. Future research should address these limitations to further strengthen the applicability of texture analysis in assessing TMJ degeneration.

5. Conclusions

With the aid of TA, CT can identify changes in the condylar medullary bone for more subtle processes, such as flattening, whereas MRI allows the identification of more advanced processes that compromise cortical integrity, such as erosion. Therefore, TA can serve as a supplement to morphological MRI and CT to enhance the identification of minor cartilage changes and as an aid in the initial treatment at an early stage.

Author Contributions

Conceptualization, A.L.F.C., C.M.O. and S.L.P.d.C.L.; investigation, C.M.O., E.F. and A.L.F.R.; formal analysis, D.L.G.L., L.A.P.d.O., B.M.d.A. and C.M.O.; data curation, E.F., A.d.O.L., K.O. and C.M.O.; methodology, A.L.F.C., A.L.F.R. and C.M.O.; project administration, A.d.O.L.; supervision, A.L.F.C. and S.L.P.d.C.L.; writing—original draft, A.L.F.C., C.M.O., S.L.P.d.C.L. and K.O.; writing—review and editing, A.L.F.C., C.M.O., E.F., D.L.G.L., L.A.P.d.O., B.M.d.A., A.d.O.L., K.O. and S.L.P.d.C.L. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

This study has been approved by the institutional review board (Ethical Committee of the School of Dentistry, University of São Paulo/USP); no. 56631222.9.0000.0075, date of approval: May 2nd 2020.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Representative sagittal images showing the central section of the condyle and the manual selection of the ROI to extract texture and image parameters from CT (A) and MRI images (B).

Figure 2. Flowchart of the methodological pipeline for the texture analysis of temporomandibular joints (TMJs) using CT and MRI images.

Figure 3. Comparison of patients with and without flattening on the right and left sides for CT and MRI images and texture parameters (angular second moment [AngScMom], contrast, correlation [Correlat], entropy, inverse difference moment [InvDfMom], sum of entropy [SumEntrp], sum of squares [SumOfSqs]). * Significant p-values.

Figure 4. Comparison of patients with and without erosion on the right and left sides for CT and MRI images and texture parameters (angular second moment [AngScMom], contrast, correlation [Correlat], entropy, inverse difference moment [InvDfMom], sum of entropy [SumEntrp], sum of squares [SumOfSqs]). * Significant p-values.

Table 1

Number of patients diagnosed for each pathology.

Pathology	Side	Number of Diagnosed Patients	Observation
Flattening	Right	22	Assessment by side
Flattening	Left	20	Assessment by side
Erosion	Right	6	Assessment by side
Erosion	Left	7	Assessment by side

Table 2

Statistically significant differences in texture parameters between groups with and without TMJ pathologies (flattening and erosion) for CT and MRI images.

Group	Texture Parameter	Position	p-Value
Flattening on the right side: CT	Higher Correlation	S(1,0)	0.005
	Higher Correlation	S(1,−1)	0.007
	Higher Correlation	S(2,0)	0.001
	Higher Correlation	S(2,−2)	0.003
	Higher Correlation	S(3,0)	0.004
	Higher Correlation	S(2,−2)	0.010
Flattening on the left side: CT	Higher Correlation	S(1,0)	0.025
	Higher Correlation	S(2,0)	0.012
	Higher Correlation	S(3,0)	0.004
	Lower Correlation	S(0,3)	0.028
Flattening on the left side: MRI	Higher Correlation	S(3,3)	0.017
Erosion on the right side: CT	Higher Contrast	All	0.018
	Higher SumOfSqs	All	0.045
	Lower Correlation	S(2,2)	0.048
Erosion on the left side: CT	Higher Correlation	S(0,1)	0.003
	Higher Correlation	S(0,2)	0.001
	Higher Correlation	S(0,3)	0.001
Erosion on the left side: MRI	Lower AngScMom	All	0.020
	Higher Contrast	All	0.031
	Higher SumOfSqs	All	0.027
	Lower InvDfMom	All	0.025
	Higher SumEntrp	All	0.027
	Higher Entropy	All	0.023

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/app14167020/s1, Table S1: The texture features computed by using MaZda software.

References

1. Valladares Neto JEC, B.M.; Guedes, O.A.; Porto, O.C.L.; Pécora, J.D. Prevalence of degenerative disease in temporomandibular disorder patients with disc displacement: A systematic review and meta-analysis. Dent. Press J. Orthod.; 2010; 15, pp. 172-181.

2. Widmalm, S.E.; Westesson, P.L.; Kim, I.K.; Pereira, F.J., Jr.; Lundh, H.; Tasaki, M.M. Temporomandibular joint pathosis related to sex, age, and dentition in autopsy material. Oral Surg. Oral Med. Oral Pathol.; 1994; 78, pp. 416-425. [DOI: https://dx.doi.org/10.1016/0030-4220(94)90031-0] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/7800370]

3. Honey, O.B.; Scarfe, W.C.; Hilgers, M.J.; Klueber, K.; Silveira, A.M.; Haskell, B.S.; Farman, A.G. Accuracy of cone-beam computed tomography imaging of the temporomandibular joint: Comparisons with panoramic radiology and linear tomography. Am. J. Orthod. Dentofac. Orthop.; 2007; 132, pp. 429-438. [DOI: https://dx.doi.org/10.1016/j.ajodo.2005.10.032] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17920495]

4. Arsan, B.; Kose, T.E.; Cene, E.; Ozcan, I. Assessment of the trabecular structure of mandibular condyles in patients with temporomandibular disorders using fractal analysis. Oral Surg. Oral Med. Oral Pathol. Oral Radiol.; 2017; 123, pp. 382-391. [DOI: https://dx.doi.org/10.1016/j.oooo.2016.11.005] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28110941]

5. Chen, J.; Sobue, T.; Utreja, A.; Kalajzic, Z.; Xu, M.; Kilts, T.; Young, M.; Wadhwa, S. Sex differences in chondrocyte maturation in the mandibular condyle from a decreased occlusal loading model. Calcif. Tissue Int.; 2011; 89, pp. 123-129. [DOI: https://dx.doi.org/10.1007/s00223-011-9498-9] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21597908]

6. Seo, B.Y.; Huh, K.H.; An, J.S.; Chang, M.S.; Ahn, S.J. Relationship of computed tomography-verified degenerative condylar morphology with temporomandibular joint disk displacement and sex. Oral Surg. Oral Med. Oral Pathol. Oral Radiol.; 2021; 132, pp. 93-103. [DOI: https://dx.doi.org/10.1016/j.oooo.2021.03.021] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34006488]

7. Song, H.; Lee, J.Y.; Huh, K.H.; Park, J.W. Long-term Changes of Temporomandibular Joint Osteoarthritis on Computed Tomography. Sci. Rep.; 2020; 10, 6731. [DOI: https://dx.doi.org/10.1038/s41598-020-63493-8] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32317672]

8. Olate, S.; Ravelo, V.; Alister, J.P.; Netto, H.D.; Haidar, Z.S.; Sacco, R. Early Treatment of Unilateral Condylar Hyperplasia in Adolescents: Preliminary Results. J. Clin. Med.; 2023; 12, 3408. [DOI: https://dx.doi.org/10.3390/jcm12103408] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37240513]

9. Vajda, D.; Chen, V.; Shailendrasinh, V.; Borg, J.; Freymiller, E. Severe skeletal deformity following post-traumatic condylar resorption: A case report. Oral Maxillofac. Surg. Cases; 2021; 7, 100227. [DOI: https://dx.doi.org/10.1016/j.omsc.2021.100227]

10. Saeed, N.R.; Gerber, B. Management of the secondary growth defect. J. Oral Biol. Craniofacial Res.; 2022; 12, pp. 833-837. [DOI: https://dx.doi.org/10.1016/j.jobcr.2022.08.006]

11. Crerand, C.E.; Sarwer, D.B.; Kazak, A.E.; Clarke, A.; Rumsey, N. Body Image and Quality of Life in Adolescents With Craniofacial Conditions. Cleft Palate Craniofacial J.; 2017; 54, pp. 2-12. [DOI: https://dx.doi.org/10.1597/15-167] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26751907]

12. Shoohanizad, E.; Garajei, A.; Enamzadeh, A.; Yari, A. Nonsurgical management of temporomandibular joint autoimmune disorders. AIMS Public Health; 2019; 6, pp. 554-567. [DOI: https://dx.doi.org/10.3934/publichealth.2019.4.554] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31909075]

13. Talmaceanu, D.; Lenghel, L.M.; Bolog, N.; Hedesiu, M.; Buduru, S.; Rotar, H.; Baciut, M.; Baciut, G. Imaging modalities for temporomandibular joint disorders: An update. Clujul Med.; 2018; 91, pp. 280-287. [DOI: https://dx.doi.org/10.15386/cjmed-970] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30093805]

14. Emshoff, R.; Brandlmaier, I.; Bertram, S.; Rudisch, A. Relative odds of temporomandibular joint pain as a function of magnetic resonance imaging findings of internal derangement, osteoarthrosis, effusion, and bone marrow edema. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod.; 2003; 95, pp. 437-445. [DOI: https://dx.doi.org/10.1067/moe.2003.95] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12686927]

15. Gillies, R.J.; Kinahan, P.E.; Hricak, H. Radiomics: Images Are More than Pictures, They Are Data. Radiology; 2016; 278, pp. 563-577. [DOI: https://dx.doi.org/10.1148/radiol.2015151169]

16. An, P.; Zhang, J.; Li, M.; Duan, P.; He, Z.; Wang, Z.; Feng, G.; Guo, H.; Li, X.; Qin, P. Clinical Data-CT Radiomics-Based Model for Predicting Prognosis of Patients with Gastrointestinal Pancreatic Neuroendocrine Neoplasms (GP-NENs). Comput. Math. Methods Med.; 2022; 2022, 4186305. [DOI: https://dx.doi.org/10.1155/2022/4186305] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36035279]

17. Samei, E.; Badano, A.; Chakraborty, D.; Compton, K.; Cornelius, C.; Corrigan, K.; Flynn, M.J.; Hemminger, B.; Hangiandreou, N.; Johnson, J. et al. Assessment of display performance for medical imaging systems: Executive summary of AAPM TG18 report. Med. Phys.; 2005; 32, pp. 1205-1225. [DOI: https://dx.doi.org/10.1118/1.1861159] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15895604]

18. Castellano, G.; Bonilha, L.; Li, L.M.; Cendes, F. Texture analysis of medical images. Clin. Radiol.; 2004; 59, pp. 1061-1069. [DOI: https://dx.doi.org/10.1016/j.crad.2004.07.008] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15556588]

19. de Oliveira, M.S.; Betting, L.E.; Mory, S.B.; Cendes, F.; Castellano, G. Texture analysis of magnetic resonance images of patients with juvenile myoclonic epilepsy. Epilepsy Behav.; 2013; 27, pp. 22-28. [DOI: https://dx.doi.org/10.1016/j.yebeh.2012.12.009]

20. de Oliveira, M.S.; D’Abreu, A.; Franca, M.C., Jr.; Lopes-Cendes, I.; Cendes, F.; Castellano, G. MRI-texture analysis of corpus callosum, thalamus, putamen, and caudate in Machado-Joseph disease. J. Neuroimaging; 2012; 22, pp. 46-52. [DOI: https://dx.doi.org/10.1111/j.1552-6569.2010.00553.x]

21. De Rosa, C.S.; Bergamini, M.L.; Palmieri, M.; Sarmento, D.J.S.; de Carvalho, M.O.; Ricardo, A.L.F.; Hasseus, B.; Jonasson, P.; Braz-Silva, P.H.; Ferreira Costa, A.L. Differentiation of periapical granuloma from radicular cyst using cone beam computed tomography images texture analysis. Heliyon; 2020; 6, e05194. [DOI: https://dx.doi.org/10.1016/j.heliyon.2020.e05194] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33088959]

22. Raja, J.V.; Khan, M.; Ramachandra, V.K.; Al-Kadi, O. Texture analysis of CT images in the characterization of oral cancers involving buccal mucosa. Dentomaxillofacial Radiol.; 2012; 41, pp. 475-480. [DOI: https://dx.doi.org/10.1259/dmfr/83345935] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22241875]

23. Feng, J.; Li, C. Editorial: Incorporation of texture analysis in diagnosing and characterizing cancer. Front. Oncol.; 2023; 13, 1224644. [DOI: https://dx.doi.org/10.3389/fonc.2023.1224644] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37346067]

24. Gomes, J.P.P.; Ogawa, C.M.; Silveira, R.V.; Castellano, G.; De Rosa, C.S.; Yasuda, C.L.; Rocha, A.C.; Hasseus, B.; Orhan, K.; Braz-Silva, P.H. et al. Magnetic resonance imaging texture analysis to differentiate ameloblastoma from odontogenic keratocyst. Sci. Rep.; 2022; 12, 20047. [DOI: https://dx.doi.org/10.1038/s41598-022-20802-7]

25. Oliveira, M.S.; Fernandes, P.T.; Avelar, W.M.; Santos, S.L.; Castellano, G.; Li, L.M. Texture analysis of computed tomography images of acute ischemic stroke patients. Braz. J. Med. Biol. Res.; 2009; 42, pp. 1076-1079. [DOI: https://dx.doi.org/10.1590/S0100-879X2009005000034] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19820884]

26. Santos, T.A.; Maistro, C.E.; Silva, C.B.; Oliveira, M.S.; Franca, M.C., Jr.; Castellano, G. MRI Texture Analysis Reveals Bulbar Abnormalities in Friedreich Ataxia. AJNR Am. J. Neuroradiol.; 2015; 36, pp. 2214-2218. [DOI: https://dx.doi.org/10.3174/ajnr.A4455] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26359147]

27. Goncalves, B.C.; de Araujo, E.C.; Nussi, A.D.; Bechara, N.; Sarmento, D.; Oliveira, M.S.; Santamaria, M.P.; Costa, A.L.F.; Lopes, S. Texture analysis of cone-beam computed tomography images assists the detection of furcal lesion. J. Periodontol.; 2020; 91, pp. 1159-1166. [DOI: https://dx.doi.org/10.1002/JPER.19-0477] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32003465]

28. Varghese, B.A.; Cen, S.Y.; Hwang, D.H.; Duddalwar, V.A. Texture Analysis of Imaging: What Radiologists Need to Know. AJR Am. J. Roentgenol.; 2019; 212, pp. 520-528. [DOI: https://dx.doi.org/10.2214/AJR.18.20624] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30645163]

29. Costa, A.L.F.; de Souza Carreira, B.; Fardim, K.A.C.; Nussi, A.D.; da Silva Lima, V.C.; Miguel, M.M.V.; Jardini, M.A.N.; Santamaria, M.P.; de Castro Lopes, S.L.P. Texture analysis of cone beam computed tomography images reveals dental implant stability. Int. J. Oral Maxillofac. Surg.; 2021; 50, pp. 1609-1616. [DOI: https://dx.doi.org/10.1016/j.ijom.2021.04.009]

30. Burr, D.B.; Gallant, M.A. Bone remodelling in osteoarthritis. Nat. Rev. Rheumatol.; 2012; 8, pp. 665-673. [DOI: https://dx.doi.org/10.1038/nrrheum.2012.130]

31. Auger, J.D.; Frings, N.; Wu, Y.; Marty, A.G.; Morgan, E.F. Trabecular Architecture and Mechanical Heterogeneity Effects on Vertebral Body Strength. Curr. Osteoporos. Rep.; 2020; 18, pp. 716-726. [DOI: https://dx.doi.org/10.1007/s11914-020-00640-0]

32. Wegrzyn, J.; Roux, J.P.; Arlot, M.E.; Boutroy, S.; Vilayphiou, N.; Guyen, O.; Delmas, P.D.; Chapurlat, R.; Bouxsein, M.L. Role of trabecular microarchitecture and its heterogeneity parameters in the mechanical behavior of ex vivo human L3 vertebrae. J. Bone Miner. Res.; 2010; 25, pp. 2324-2331. [DOI: https://dx.doi.org/10.1002/jbmr.164] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20564249]

33. Brandi, M.L. Microarchitecture, the key to bone quality. Rheumatology; 2009; 48, (Suppl. S4), pp. iv3-iv8. [DOI: https://dx.doi.org/10.1093/rheumatology/kep273] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19783591]

34. Cardoneanu, A.; Macovei, L.A.; Burlui, A.M.; Mihai, I.R.; Bratoiu, I.; Rezus, I.I.; Richter, P.; Tamba, B.I.; Rezus, E. Temporomandibular Joint Osteoarthritis: Pathogenic Mechanisms Involving the Cartilage and Subchondral Bone, and Potential Therapeutic Strategies for Joint Regeneration. Int. J. Mol. Sci.; 2022; 24, 171. [DOI: https://dx.doi.org/10.3390/ijms24010171] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36613615]

35. Roberts, W.E.; Stocum, D.L. Part II: Temporomandibular Joint (TMJ)-Regeneration, Degeneration, and Adaptation. Curr. Osteoporos. Rep.; 2018; 16, pp. 369-379. [DOI: https://dx.doi.org/10.1007/s11914-018-0462-8] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29943316]

36. Nurzynska, K.; Piórkowski, A.; Strzelecki, M.; Kociołek, M.; Banyś, R.P.; Obuchowicz, R. Differentiating age and sex in vertebral body CT scans—Texture analysis versus deep learning approach. Biocybern. Biomed. Eng.; 2024; 44, pp. 20-30. [DOI: https://dx.doi.org/10.1016/j.bbe.2023.11.002]

37. Obuchowicz, R.; Strzelecki, M.; Piorkowski, A. Clinical Applications of Artificial Intelligence in Medical Imaging and Image Processing-A Review. Cancers; 2024; 16, 1870. [DOI: https://dx.doi.org/10.3390/cancers16101870] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38791949]

38. Girondi, C.M.; de Castro Lopes, S.L.P.; Ogawa, C.M.; Braz-Silva, P.H.; Costa, A.L.F. Texture Analysis of Temporomandibular Joint Disc Changes Associated with Effusion Using Magnetic Resonance Images. Dent. J.; 2024; 12, 82. [DOI: https://dx.doi.org/10.3390/dj12030082]

39. Ricardo, A.L.F.; da Silva, G.A.; Ogawa, C.M.; Nussi, A.D.; De Rosa, C.S.; Martins, J.S.; de Castro Lopes, S.L.P.; Appenzeller, S.; Braz-Silva, P.H.; Costa, A.L.F. Magnetic resonance imaging texture analysis for quantitative evaluation of the mandibular condyle in juvenile idiopathic arthritis. Oral Radiol.; 2023; 39, pp. 329-340. [DOI: https://dx.doi.org/10.1007/s11282-022-00641-y]

40. Nussi, A.D.; de Castro Lopes, S.L.P.; De Rosa, C.S.; Gomes, J.P.P.; Ogawa, C.M.; Braz-Silva, P.H.; Costa, A.L.F. In vivo study of cone beam computed tomography texture analysis of mandibular condyle and its correlation with gender and age. Oral Radiol.; 2023; 39, pp. 191-197. [DOI: https://dx.doi.org/10.1007/s11282-022-00620-3]