DEEP LEARNING-BASED MANDIBULAR MOLARS DETECTION AND CLASSIFICATION OF FURCATION INVOLVEMENT

Abstract

The present study aims to investigate the potential of deep learning methodologies in enhancing diagnostic accuracy and healthcare outcomes in dentistry. Specifically, the study explores the effectiveness of convolutional neural networks (CNNs) in detecting dental abnormalities and distinguishing between healthy teeth and those exhibiting signs of furcation involvement (FI). The research findings suggest that CNNs outperform traditional machine learning models in classifying dental imaging data, particularly in detecting furcation involvement. This highlights the potential of deep learning algorithms in medical image analysis tasks. The study also presents a developed algorithm utilizing the Faster R-CNN model, demonstrating promising capabilities in accurately detecting individual teeth on radiographs and streamlining diagnostic procedures. The developed analysis tool offers users an interactive interface to select regions of interest and obtain classification results with ease and precision. Overall, this research highlights the valuable role of artificial intelligence in assisting clinicians with early disease detection and treatment planning in dentistry, which can improve patient care and outcomes in dental healthcare.