Research map / Graph learning / Graph self-supervised learning
GIIM: Graph-based Learning of Inter- and Intra-view Dependencies for Multi-view Medical Image Diagnosis
Tran Bao Sam, Hung Vu, Dao Trung Kien, Tran Dat Dang, Van Ha Tang, Steven Truong
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 |
|---|---|---|---|
| model inter- and intra-view dependencies with a heterogeneous graph (GIIM) instead of NN/ML/attention fusion | multi-view ConvNeXt features · Liver tumor CT test | accuracy (%): 75.45 -> 78.20 | Improved |
| GIIM vs best fusion baseline | multi-view ConvNeXt features · BreastDM | accuracy (%): 85.1 -> 87.23 | Improved |
| GIIM vs best fusion baseline on AUC | multi-view ConvNeXt features · BreastDM | AUC (%): 87.35 -> 89.02 | Mixed |
| GIIM vs NN-based on VinDr-Mammo AUC | multi-view ConvNeXt features · VinDr-Mammo test | AUC (%): 82.21 -> 82.54 | No clear effect |
| use multiple views instead of a single view | ConvNeXt classifier · Liver tumor CT test | accuracy (%): 68.64 -> 78.20 | Improved |
| constant missing-view vector vs other imputations on missing-view test (ablation) | GIIM · Liver, missing-view test | accuracy (%): 72.27 (constant) vs 73.64 (learnable), 74.31 (RAG), 70.91 (covariance) | Mixed |
| RAG or covariance imputation on full-view test (ablation) | GIIM · VinDr-Mammo, full-view test | accuracy (%): 71.17 -> 71.17 | No clear effect |
| learnable missing vector with high missing rate in training (eta=1.0) (ablation) | GIIM · VinDr-Mammo, full-view test | accuracy (%): 71.17 -> 60.74 | Hurt |
| covariance imputation at eta=0.2 (ablation) | GIIM · Liver, full-view test | accuracy (%): 78.20 -> 78.41 | Mixed |
| test with a missing view instead of full views | GIIM · Liver and VinDr-Mammo | accuracy (%): 78.20 -> 72.27 | Hurt |
Closest papers by meaning
- M3CoTBench: Benchmark Chain-of-Thought of MLLMs in Medical Image Understanding (ICLR 2026 · similarity 0.84)
- TumorChain: Interleaved Multimodal Chain-of-Thought Reasoning for Traceable Clinical Tumor Analysis (ICLR 2026 · similarity 0.84)
- Large-scale and Fine-grained Vision-language Pre-training for Enhanced CT Image Understanding (ICLR 2025 · similarity 0.83)
- Gastric-X: A Multimodal Multi-Phase Benchmark Dataset for Advancing Vision-Language Models in Gastric Cancer Analysis (CVPR 2026 · similarity 0.83)
- DDx-TRACE: A Benchmark for Medical Diagnostic Trajectories in VLMs (NeurIPS 2026 · similarity 0.83)
- CXReasonBench: A Benchmark for Evaluating Structured Diagnostic Reasoning in Chest X-rays (NeurIPS 2025 · similarity 0.82)
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.
