Extraction of 3D trajectories of mandibular condyles from 2D real-time MRI

Karyna Isaieva,Justine Leclère,Guillaume Paillart,Guillaume Drouot,Jacques Felblinger,Xavier Dubernard,Pierre-André Vuissoz

2024-06-21

Abstract:Computing the trajectories of mandibular condyles directly from MRI could provide a comprehensive examination, allowing for the extraction of both anatomical and kinematic details. This study aimed to investigate the feasibility of extracting 3D condylar trajectories from 2D real-time MRI and to assess their precision.Twenty healthy subjects underwent real-time MRI while opening and closing their jaws. One axial and two sagittal slices were segmented using a U-Net-based algorithm. The centers of mass of the resulting masks were projected onto the coordinate system based on anatomical markers and temporally adjusted using a common projection. The quality of the computed trajectories was evaluated using metrics designed to estimate movement reproducibility, head motion, and slice placement symmetry.The segmentation of the axial slices demonstrated good-to-excellent quality; however, the segmentation of the sagittal slices required some fine-tuning. The movement reproducibility was acceptable for most cases; nevertheless, head motion displaced the trajectories by 1 mm on average. The difference in the superior-inferior coordinate of the condyles in the closed jaw position was 1.7 mm on average.Despite limitations in precision, real-time MRI enables the extraction of condylar trajectories with sufficient accuracy for evaluating clinically relevant parameters such as condyle displacement, trajectories aspect, and symmetry.

Image and Video Processing,Artificial Intelligence