Abstract

This study aims to develop a deep learning framework for the automatic detection of the position relationship between the mandibular third molar (M3) and the mandibular canal (MC) on panoramic radiographs (PRs), to assist doctors in assessing and planning appropriate surgical interventions.