AI-Powered Classification of Root Canal Irrigation Efficiency Based on Image Analysis

Abstract

Root canal irrigation plays a vital role in the endodontic treatment success, since it guarantees the 
elimination of debris, microorganisms, and biofilm of the complex root canal system. Conventional 
irrigation efficiency assessment techniques tend to be subjective, laborious and low reproducibility. In 
this research, the authors explain how artificial intelligence (AI) and image analysis can be used to 
categorize the efficacy of root canal irrigation. Images of treated canals of high resolution were obtained 
and preprocessed to extract features. A convolutional neural network (CNN)-based model was trained to 
read morphological patterns and categorize the results of the irrigation as efficient or inefficient. The 
accuracy of the AI model during the classification process was rather high in the comparison with expert 
ratings, which indicates the possibility of objective and automated assessment. The results indicate that 
AI-based image recognition can be an effective supplement to an endodontic study and clinical practice 
and provide a stable and quick assessment of irrigation procedures. Additional future studies are advised 
to be conducted on a larger scale, in real time, and to be validated in clinical settings to improve reliability 
and become part of routine care.