Inferior Alveolar Nerve Segmentation in CBCT images using Connectivity-Based Selective Re-training
This work addresses the need for accurate nerve detection in dental surgery to prevent nerve damage, representing an incremental improvement in a domain-specific application.
The paper tackled the problem of Inferior Alveolar Nerve segmentation in CBCT images by transforming a mixed supervised problem into a semi-supervised one using a connectivity-based selective re-training framework, achieving a dice similarity coefficient of 0.7956 and winning the ToothFairy2023 Challenge.
Inferior Alveolar Nerve (IAN) canal detection in CBCT is an important step in many dental and maxillofacial surgery applications to prevent irreversible damage to the nerve during the procedure.The ToothFairy2023 Challenge aims to establish a 3D maxillofacial dataset consisting of all sparse labels and partial dense labels, and improve the ability of automatic IAN segmentation. In this work, in order to avoid the negative impact brought by sparse labeling, we transform the mixed supervised problem into a semi-supervised problem. Inspired by self-training via pseudo labeling, we propose a selective re-training framework based on IAN connectivity. Our method is quantitatively evaluated on the ToothFairy verification cases, achieving the dice similarity coefficient (DSC) of 0.7956, and 95\% hausdorff distance (HD95) of 4.4905, and wining the champion in the competition. Code is available at https://github.com/GaryNico517/SSL-IAN-Retraining.