A NATOMY -G UIDED PARALLEL B OTTLENECK T RANSFORMER N ETWORK FOR AUTOMATED E VALUATION OF ROOT C ANAL T HERAPY Yunxiang Li*, Lingling Sun, Neng Xia, Ruizi Peng, Kai Tang Hangzhou Dianzi University Hangzhou, China {li1124325213, sunll, amoreyo, prince75, 19061129}@hdu.edu.cn Guodong Zeng* University of Bern Swiss guodong.zeng@sitem.unibe.ch Yifan Zhang, Qisi Lian West China Hospital of Stomatology, Sichuan University Chengdu, China zhangyifan@hzyk.com.cn, qisiscu@163.com Jun Wang Shanghai Jiao Tong University 12474 Shanghai, China wjcy19870122@sjtu.edu.cn Qianni Zhang Queen Mary University of London London, UK qianni.zhang@qmul.ac.uk Qun Jin Waseda University 13148 Shinjuku-ku, Tokyo, Japan jin@waseda.jp Yaqi Wang † Communication University of Zhejiang 92254 Hangzhou, Zhejiang, China wangyaqi@cuz.edu.cn Shuai Wang † Shandong University Weihai, China shuaiwang@sdu.edu.cn ABSTRACT Objective: Accurate evaluation of the root canal filling result in X-ray image is a significant step for the root canal therapy, which is based on the relative position between the apical area boundary of tooth root and the top of filled gutta-percha in root canal as well as the shape of the tooth root and so on to classify the result as correct-filling, under-filling or over-filling. Methods: We propose a novel anatomy-guided Transformer diagnosis network. For obtaining accurate anatomy-guided features, a polynomial curve fitting segmentation is proposed to segment the fuzzy boundary. And a Parallel Bottleneck Transformer network (PBT-Net) is introduced as the classification network for the final evaluation. Results, and conclusion: Our numerical experiments show that our anatomy-guided PBT-Net improves the accuracy from 40% to 85% relative to the baseline classification network. Comparing with the SOTA segmentation network indicates that the ASD is significantly reduced by 30.3% through our fitting segmentation. Significance: Polynomial curve fitting segmentation has a great segmentation effect for extremely fuzzy boundaries. The prior knowledge guided classification network is suitable for the evaluation of root canal therapy greatly. And the new proposed Parallel Bottleneck Transformer for realizing self-attention is general in design, facilitating a broad use in most backbone networks. Keywords Root Canal Therapy · Transformer · Classification · Segmentation · X-ray arXiv:2105.00381v1 [cs.CV] 2 May 2021