(TC11) Enhanced Diagnostic Imaging Using Artificial Intelligence in Endodontics

Accurate diagnosis is essential for effective treatment of endodontic patients. There are currently areas in endodontic differential diagnosis that are limited and require invasive procedures. For example, the differentiation of periapical granulomas and cysts, especially in cases with persistent disease, may assist with treatment planning. Artificial intelligence (AI), when applied to cone-beam computed tomography (CBCT), enhances diagnostic precision by analyzing 3D radiographs through convolutional neural networks (CNNs). These networks detect subtle radiographic differences, providing clinicians with a prioritized list of potential diagnoses and reducing interobserver variability. Beyond granuloma-cyst differentiation, AI offers new opportunities to assess outcomes of regenerative endodontic therapy (imaging revitalized tissue), monitor the healing of large apical lesions, detect root fractures and cracks, and distinguish between vital and necrotic pulp tissue.

This table clinic will explore the scientific principles behind AI-powered CBCT analysis, detailing the processes of data preprocessing, model training, and the generation of differential diagnoses. Attendees will gain insight into how AI algorithms learn from diverse datasets, ensuring robust and reliable clinical support. Additionally, the presentation will address critical considerations such as dataset validation and the implementation of privacy-preserving technologies like federated learning, which safeguards patient data while enabling collaborative AI model development.

By understanding the scientific mechanisms and practical applications of AI in CBCT analysis, attendees will appreciate how this technology enhances diagnostic workflows, improves decision-making, and ultimately elevates patient care in endodontics. This presentation aims to demonstrate the transformative potential of AI-driven diagnostic tools in clinical practice.