Evaluating Artificial Intelligence in Full-Arch CBCT Caries Detection: A Comparative Analysis with Clinical...

Area of Science:

• Dentistry
• Artificial Intelligence
• Medical Imaging

Background:

• Dental caries detection is crucial for timely intervention.
• Cone-beam computed tomography (CBCT) is a valuable imaging tool in dentistry.
• AI-powered diagnostic systems offer potential advancements in dental diagnostics.

Purpose of the Study:

• To compare the diagnostic accuracy of the Diagnocat AI system against clinical assessments by dentists for detecting carious lesions.
• To evaluate the performance of an AI system in analyzing CBCT scans for caries detection.

Main Methods:

• Analysis of CBCT data for carious lesion detection.
• Comparison of AI diagnoses with assessments from three experienced dentists under single-blind conditions.
• Evaluation of diagnostic agreement based on tooth location, patient age, and gender.

Main Results:

• Variable agreement between the Diagnocat system and dentists, influenced by tooth type, patient age, and gender.
• Lower agreement observed for premolars due to complex morphology.
• Higher diagnostic accuracy noted for molars and incisors, especially in younger patients.

Conclusions:

• The Diagnocat system shows promise as a supportive and screening tool in dental practice for preliminary imaging evaluation.
• Integrating multiple diagnostic modalities like intraoral scans and photographic documentation may enhance diagnostic precision, particularly for early-stage lesions.
• Further validation and refinement of AI algorithms are needed for comprehensive caries detection.