Huazheng Zhu, Zicheng Nie, Yaping Liu, Fukang Yang, Liang Chen, Yuanyuan Jia
Assessment of root canal treatment quality is central to improving endodontic outcomes, yet interpretation of periapical radiographs remains susceptible to observer variability, fatigue, and anatomical complexity. To provide a more objective and efficient approach to post-endodontic evaluation, we developed MS-WEFNet, an enhanced YOLOv10n-based object detection framework for automatic assessment of root canal filling quality. The network incorporates a Multi-Scale Edge Information Enhancement module to reinforce fine anatomical boundaries and a Wavelet-guided Contrastive Feature Aggregation module to improve multi-scale structural representation. MS-WEFNet was evaluated on the T2k dataset, which comprised 2564 periapical radiographs containing 6772 annotated teeth. Experienced endodontists assigned the teeth to three clinical categories: 4032 non-target teeth, 1417 qualified teeth, and 1323 teeth requiring warning. The warning category encompassed underfilling, overfilling, and instrument separation. The dataset was partitioned into training, validation, and testing subsets at a ratio of 7:2:1. The proposed model achieved an mAP 50 of 0.902 ± 0.003 and an mAP 50 - 95 of 0.817 ± 0.005, representing improvements of 4.0% and 4.3%, respectively, over the YOLOv10n baseline. Additional clinical validation was further conducted on an independently collected real-world clinical cohort comprising 198 cases and 392 teeth. The assessments of three experienced endodontists served as the reference standard, against which MS-WEFNet yielded an overall false detection rate of 8.7%. Given that the CAD system is intended to support comprehensive post-treatment quality monitoring and that all outputs remain subject to clinician review, this false detection rate may be acceptable for preliminary screening in routine clinical practice. These results support the potential use of MS-WEFNet as a computer-aided tool for root canal filling quality assessment.