Diagnostic Accuracy of a Deep Learning Framework for Automated Evaluation of Root Canal Obturation Quality From Periapical Radiographs

Diagnostic Accuracy of a Deep Learning Framework for Automated Classification, Quantitative Assessment and Comprehensive Evaluation of Root Canal Obturation Quality From Periapical Radiographs

This study aims to develop and evaluate an artificial intelligence (AI)-based system that can automatically assess the quality of root canal fillings using dental X-ray images. The AI system will analyze important features of the filling, including its length, uniformity, and shape, and classify the treatment quality as acceptable or needing improvement.

The study will use previously collected, anonymized dental X-ray images of teeth that have received root canal treatment. Experienced dental specialists will evaluate these images to provide a reference standard, which will be compared with the AI system's results.

The goal of this research is to determine whether AI can provide a reliable and consistent method for evaluating root canal treatment outcomes. In the future, such technology may help dentists make more accurate decisions, improve treatment evaluation, and contribute to better patient care.

Periapical radiographs of teeth with completed root canal treatment from patients Aged between 18 and 60 years will be included, provided they exhibit satisfactory image quality characterized by adequate sharpness, contrast, and minimal noise, allowing clear visualization of the root canal filling and apical region. The radiographs must enable accurate assessment of obturation quality, including filling length, homogeneity, and taper. Both single-rooted and multi-rooted teeth will be considered to ensure adequate anatomical representation. Radiographs with poor image quality, significant distortion, metallic artifacts, post-core restorations, root resorption, fractures, or incomplete visualization of the apex will be excluded to ensure reliable analysis.