Ribera, Nina TubauDe Dumast, PriscilleYatabe, MariliaRuellas, AntonioIoshida, MarcosPaniagua, BeatrizStyner, MartinGonçalves, João Roberto [UNESP]Bianchi, Jonas [UNESP]Cevidanes, LuciaPrieto, Juan-Carlos2019-10-062019-10-062019-01-01Progress in Biomedical Optics and Imaging - Proceedings of SPIE, v. 10950.1605-7422http://hdl.handle.net/11449/190455We developed a deep learning neural network, the Shape Variation Analyzer (SVA), that allows disease staging of bony changes in temporomandibular joint (TMJ) osteoarthritis (OA). The sample was composed of 259 TMJ CBCT scans for the training set and 34 for the testing dataset. The 3D meshes had been previously classified in 6 groups by 2 expert clinicians. We improved the robustness of the training data using data augmentation, SMOTE, to alleviate over-fitting and to balance classes. We combined geometrical features and a shape descriptor, heat kernel signature, to describe every shape. The results were compared to nine different supervised machine learning algorithms. The deep learning neural network was the most accurate for classification of TMJ OA. In conclusion, SVA is a 3D Slicer extension that classifies pathology of the temporomandibular joint osteoarthritis cases based on 3D morphology.engClassificationDeep LearningNeural NetworkOsteoarthritisTemporomandibular Joint DisordersTrabalho apresentado em evento10.1117/12.2506018Acesso aberto2-s2.0-85068192586