CED-IADR/NOF Oral Health Research Congress, Geneve, Switzerland, 12 - 14 September 2024, pp.1
Objectives The purpose of this study was to propose a machine-learning model and
assess its ability to classify temporomandibular joint (TMJ) disk displacements on
magnetic resonance (MR) T1-W and PD-W images.
Methods This retrospective cohort study included 180 TMJs from 90 patients with TMJ
signs and symptoms. A Radiomics platform was used to extract (Huiying Medical
Technology Co., Ltd, China) imaging features of TMJ pathologies, condylar bone
changes and disc displacements. Thereafter, different machine learning (ML) algorithms
and logistic regression were implemented on radiomic features for feature selection,
classification, and prediction. The following radiomic features included first-order
statistic, shape, texture, gray level co-occurrence matrix (GLCM), gray level run length
matrix (GLRLM) and gray level size zone matrix (GLSZM). Six classifiers, including logistic
regression (LR), random forest (RF), decision tree (DT), k-nearest neighbors (KNN),
XGBoost and support vector machine (SVM) were used for a model building which could
predict the TMJ pathologies. The performance of models was evaluated by sensitivity,
specificity and ROC curve. The TMJ diskdisplacements were classified as; (0) Normal,
(1) ADDwR, (2) ADDwoR.
Results A total of 90 patients and 180 TMJs (19 men and 71 women; mean age,
33.6±16.8; range between 13-79 years) were included in this study. KNN classifier was
found to be the most optimal machine learning model for prediction of TMJ pathologies.
The AUC, sensitivity, and specificity for the training set were 0.944, 0.771, 0.918 for
normal, ADDwR and ADDwoR while testing set were 0.913, 0.716, 1 for normal, ADDwR
and ADDwoR. For TMJ Disk Diplacment Large Area High Gray Level Emphasis,
firstorder_Skewness, firstorder_minumum, RootMeanSquared,
GrayLevelNonUniformity, firstorder_ Kurtosis, Long Run High Gray Level Emphasis, were
selected.
Conclusions This study has proposed a machine learning model by KNN analysis on
TMJ MR images, which can be used to TMJ disc displacements.