Development of deep learning based- automatic full mouth series (FMX) mounting program to enhance pre-clinical and clinical teaching and patient care at East Carolina University School of Dental Medicine (ECU SoDM)

An FMX consists of 18 periapical and bitewing radiographs and is the most commonly used imaging modality for diagnosing and planning treatment for dental diseases. Correctly mounting these images in the FMX template is critical, as improper mounting may lead to misdiagnosis and mistreatment. Learning to mount FMX poses challenges for dental students due to overlapping coverage in certain projections and anatomical variations. However, the limited number of instructors available to supervise students during the mounting process, combined with technical complexity of post-mounting corrections, calls for innovative and effective strategies to address this educational challenge. The purpose of this proposal is to develop a deep learning-based algorithm for automatic FMX mounting to assist in pre-clinical and clinical radiology teaching, as well as patient care, at ECU SoDM. The envisioned program will identify mounting errors and alert the operator for correction until all images are displayed correctly. This program is expected to serve as a vital supplementary tool for courses DENT8420 ORAD02 Radiology Foundation and CF03.01 Clinical Radiology, ensuring an efficient and effective learning experience for students, as well as safe, high-quality care for our patients.

Scholars @ ECU