MICCAI STS 2024 challenge: Semi-supervised instance-level tooth segmentation in panoramic X-ray and CBCT images.
Authors
Affiliations (18)
Affiliations (18)
- Innovation Center for Electronic Design Automation Technology, Hangzhou Dianzi University, Hangzhou, China.
- School of Cyberspace, Hangzhou Dianzi University, Hangzhou, China.
- Shandong University, Weihai, China.
- Innovation Center for Electronic Design Automation Technology, Hangzhou Dianzi University, Hangzhou, China. Electronic address: [email protected].
- Hangzhou Geriatric Stomatology Hospital, Hangzhou Dental Hospital Group, Hangzhou, China.
- Hangzhou Dianzi University, Hangzhou, China.
- Shenzhen University, Shenzhen, China.
- Shanghai Jiao Tong University, Shanghai, China.
- Xidian University, Xi'an, China.
- Harbin Institute of Technology, Harbin, China.
- The University of Hong Kong, Hong Kong SAR, China.
- Chinese Academy of Sciences, Chengdu, China.
- Yunnan Provincial Stomatology Hospital, Kunming, China.
- Shanghai MediWorks Precision Instruments Co, Ltd, Shanghai, China.
- Department of Stomatology, The First Affilliated Hospital, College of Medicine, Zhejiang University, Hangzhou, China.
- Queen Mary University of London, London, United Kingdom.
- School of Cyberspace, Hangzhou Dianzi University, Hangzhou, China. Electronic address: [email protected].
- University of Leicester, Leicester, United Kingdom. Electronic address: [email protected].
Abstract
Orthopantomogram (OPGs) and Cone-Beam Computed Tomography (CBCT) are vital for dentistry, but creating large datasets for automated tooth segmentation is hindered by the labor-intensive process of manual instance-level annotation. This research aimed to benchmark and advance semi-supervised learning (SSL) as a solution for this data scarcity problem. We organized the 2nd Semi-supervised Teeth Segmentation (STS 2024) Challenge at MICCAI 2024. We provided a large-scale dataset comprising over 90,000 2D images and 3D axial slices, which includes 2380 OPG images and 330 CBCT scans, all featuring detailed instance-level FDI annotations on part of the data. The challenge attracted 114 (OPG) and 106 (CBCT) registered teams. To ensure algorithmic excellence and full transparency, we rigorously evaluated the valid, open-source submissions from the top 10 (OPG) and top 5 (CBCT) teams, respectively. All successful submissions were deep learning-based SSL methods. The winning semi-supervised models demonstrated impressive performance gains over a fully-supervised nnU-Net baseline trained only on the labeled data. For the 2D OPG track, the top method improved the Instance Affinity (IA) score by over 44 percentage points. For the 3D CBCT track, the winning approach boosted the Instance Dice score by 61 percentage points. This challenge demonstrates the potential benefit benefit of SSL for complex, instance-level medical image segmentation tasks where labeled data is scarce. The most effective approaches consistently leveraged hybrid semi-supervised frameworks that combined knowledge from foundational models like SAM with multi-stage, coarse-to-fine refinement pipelines. Both the challenge dataset and the participants' submitted code have been made publicly available on GitHub (https://github.com/ricoleehduu/STS-Challenge-2024), ensuring transparency and reproducibility.