Research map / Label-efficient and robust learning / Semi-supervised learning
Prior-Aware Neural Network for Partially-Supervised Multi-Organ Segmentation
Yuyin Zhou, Zhe Li, Song Bai, Chong Wang, Xinlei Chen, Mei Han, Elliot Fishman, Alan L. Yuille
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add prior-aware (organ size KL) loss on top of naive partial supervision (ablation) | ResNet-50 with partial data A+B+C · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7705 -> 0.7833 | Improved |
| add prior-aware (organ size KL) loss on top of naive partial supervision (ablation) | ResNet-101 with partial data A+B+C · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7735 -> 0.7904 | Improved |
| add prior-aware (organ size KL) loss on top of naive partial supervision (ablation) | 3D-UNet with partial data A+B+C · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7163 -> 0.7208 | Improved |
| treat partial labels as real labels instead of unlabeled data (ablation) | ResNet-50 with partial data A+B+C · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7669 -> 0.7705 | Improved |
| add partially-labeled data with the prior-aware loss | ResNet-50 fully supervised · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7535 -> 0.7833 | Improved |
| use all three partial datasets rather than one (PaNN) (ablation) | ResNet-101 · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7748 -> 0.7904 | Improved |
| use all three partial datasets rather than one (semi-supervised) (ablation) | ResNet-101 · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7637 -> 0.7719 | Improved |
| use all three partial datasets rather than one (naive partial) (ablation) | ResNet-101 · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7684 -> 0.7735 | Improved |
| use 2D models instead of 3D models | PaNN · MICCAI 2015 multi-atlas abdomen | average Dice: 0.7208 -> 0.7833 | Improved |
| PaNN with partial dataset C (liver) on a non-target organ (ablation) | ResNet-50 · MICCAI 2015 multi-atlas abdomen | gallbladder Dice: 68.97% -> 72.26% | Improved |
| PaNN with partial dataset C (liver) on a hard organ | ResNet-50 · MICCAI 2015 multi-atlas abdomen | left adrenal gland Dice: 49.51% -> 53.63% | Improved |
| add prior-aware loss to naive partial supervision (ablation) | ResNet · New abdominal CT dataset (9 organs) | average Dice: 0.9063 -> 0.9103 | Improved |
Closest papers by meaning
- AMOS: A Large-Scale Abdominal Multi-Organ Benchmark for Versatile Medical Image Segmentation (NeurIPS 2022 · similarity 0.90)
- AbdomenAtlas-8K: Annotating 8,000 CT Volumes for Multi-Organ Segmentation in Three Weeks (NeurIPS 2023 · similarity 0.89)
- AbdomenCT-1K: Is Abdominal Organ Segmentation a Solved Problem? (TPAMI 2021 · similarity 0.89)
- CAT: Coordinating Anatomical-Textual Prompts for Multi-Organ and Tumor Segmentation (NeurIPS 2024 · similarity 0.84)
- PanTS: The Pancreatic Tumor Segmentation Dataset (NeurIPS 2025 · similarity 0.84)
- Touchstone Benchmark: Are We on the Right Way for Evaluating AI Algorithms for Medical Segmentation? (NeurIPS 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
