Automatic Detection of Mandibular Fractures in Panoramic Radiographs Using Deep Learning

Area of Science:

• Oral and Maxillofacial Surgery
• Medical Imaging
• Artificial Intelligence in Medicine

Background:

• Mandibular fractures are common in oral and maxillo-facial surgery.
• Diagnosis typically relies on panoramic radiography or cone-beam computed tomography (CBCT).
• Panoramic radiography is a simpler imaging technique compared to CBCT.

Purpose of the Study:

• To develop a deep learning system for automated mandibular fracture diagnosis using panoramic radiographs.
• To evaluate the performance of a YOLO-based deep learning model for this task.
• To enhance the diagnostic accuracy of mandibular fracture detection.

Main Methods:

• Implementation of a YOLO (You Only Look Once) deep learning model for fracture detection.
• Augmentation of panoramic radiograph images using gamma modulation, multi-bounding boxes, and luminance adaptation transforms.
• Testing the deep learning system's performance on panoramic radiographs.

Main Results:

• The YOLO-based deep learning system demonstrated superior detection performance compared to conventional methods.
• Image augmentation techniques improved the accuracy of the deep learning model.
• The proposed method offers a reliable tool for identifying mandibular fractures.

Conclusions:

• A deep learning system based on YOLO can effectively diagnose mandibular fractures from panoramic radiographs.
• The developed system shows potential to assist radiologists in double-checking diagnoses.
• This AI-driven approach enhances the efficiency and accuracy of mandibular fracture detection.