Detection and classification of dental deficiencies from panoramic dental X-rays using machine learning

Recent advancements in dental research driven by digital technologies and artificial intelligence (AI) have radically transformed diagnostic and treatment planning processes. AI, possessing human-like capabilities, plays a particularly transformative role in the field of dentistry by enabling precise analysis and personalized treatment options. In panoramic radiographs, AI emerges as a critical tool for early disease detection and treatment optimization. This thesis aims to propose an AI-based clinical decision support system in the field of dental health. To achieve this, missing teeth and the suitability of implant placement in these regions will be identified using machine learning techniques applied to panoramic dental radiographs. Accordingly, two publicly available datasets were combined and processed to create a new dataset, along with a real dataset composed of patient panoramic radiographic images obtained from Hamidiye Faculty of Dentistry, University of Health Sciences. Various machine learning techniques and hybrid architectures were applied to these datasets, and their performance results were analyzed to determine the most suitable intelligent approach to the problem. As a result of the evaluation, the proposed computer-aided diagnostic system is expected to support patient care and treatment decision-making processes in dentistry.