Research map / Detection and segmentation / CNN features
3D Graph Anatomy Geometry-Integrated Network for Pancreatic Mass Segmentation, Diagnosis, and Quantitative Patient Management
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Integrate the anatomy mesh and graph classification with segmentation | nnUNet-based pancreatic mass segmentation · NonPDAC masses | Average Dice: 0.478 -> 0.611 | Helped |
| Integrate the anatomy mesh and graph classification with segmentation | nnUNet-based pancreatic mass segmentation · NonPDAC masses | Detection rate: 69.3% -> 88.7% | Helped |
| Use all five CT phases instead of one phase (ablation) | SMCN · PDAC segmentation | Average Dice: 0.730 -> 0.738 | Helped |
| Use global classification rather than pixel voting (ablation) | SMCN · PDAC vs. nonPDAC | Accuracy: 0.913 -> 0.927 | Helped |
| Use global classification rather than vertex-based voting (ablation) | SMCN · PDAC vs. nonPDAC | Specificity: 0.891 -> 0.906 | Helped |
| Use global classification rather than pixel voting (ablation) | SMCN · Patient management | Accuracy: 0.832 -> 0.865 | Helped |
| Use global classification rather than vertex-based voting (ablation) | SMCN · Patient management | Average Recall: 0.656 -> 0.746 | Helped |
| Use semi-supervised pseudo nonPDAC masks instead of annotated nonPDAC masks (ablation) | SMCN · PDAC vs. nonPDAC detection | nonPDAC sensitivity: 0.906 -> 0.515 | Hurt |
| Use semi-supervised pseudo nonPDAC masks instead of annotated nonPDAC masks (ablation) | SMCN · PDAC detection | Detection rate: 0.986 -> 0.959 | Hurt |
| Use imaging-only automated classification | SMCN · PDAC vs. nonPDAC | Sensitivity: 0.945 vs. ~0.94 | No clear effect |
| Use imaging-only automated classification | SMCN · PDAC vs. nonPDAC | Specificity: 0.906 vs. ~0.90 | No clear effect |
| Use five-phase CT SMCN instead of the multimodality CompCyst test | Patient management · Surgery recommendation | Correct recommendation rate: 91% -> 95% | Helped |
Closest papers by meaning
- Recurrent Saliency Transformation Network: Incorporating Multi-Stage Visual Cues for Small Organ Segmentation (CVPR 2018 · similarity 0.87)
- DoDNet: Learning To Segment Multi-Organ and Tumors From Multiple Partially Labeled Datasets (CVPR 2021 · similarity 0.87)
- AMOS: A Large-Scale Abdominal Multi-Organ Benchmark for Versatile Medical Image Segmentation (NeurIPS 2022 · similarity 0.87)
- PanTS: The Pancreatic Tumor Segmentation Dataset (NeurIPS 2025 · similarity 0.86)
- Upping the Game: How 2D U-Net Skip Connections Flip 3D Segmentation (NeurIPS 2024 · similarity 0.85)
- ZePT: Zero-Shot Pan-Tumor Segmentation via Query-Disentangling and Self-Prompting (CVPR 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.