Research map / Label-efficient and robust learning / Recognition and long-tail
ChestX-ray8: Hospital-Scale Chest X-Ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases
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 |
|---|---|---|---|
| Initialize the weakly supervised classifier with ResNet-50 rather than the other tested CNNs (ablation) | Unified multi-label CNN on ChestX-ray8 · ChestX-ray8 | AUC: 0.5609 to 0.7164 for Nodule versus VGGNet-16 0.6556; ResNet-50 has the best AUC for seven of eight diseases | Helped |
| Balance positive and negative examples in the classification loss (ablation) | ResNet-50 multi-label classifier · ChestX-ray8, Cardiomegaly | AUC: 0.7262 to 0.8141 | Helped |
| Balance positive and negative examples in the classification loss (ablation) | ResNet-50 multi-label classifier · ChestX-ray8, Pneumonia | AUC: 0.5164 to 0.6333 | Helped |
| Use LSE pooling with r=10 rather than average or max pooling (ablation) | ResNet-50 multi-label classifier · ChestX-ray8 | AUC | Helped |
| Use average pooling rather than max pooling (ablation) | ResNet-50 multi-label classifier · ChestX-ray8 | AUC | No clear effect |
| Use syntactic-level negation and uncertainty detection in report label mining | Radiology report labeling · OpenI | F1-score: 0.86 to 0.90 | Helped |
| Expand the dataset label set from eight to fourteen diseases (ablation) | ResNet-50 classification benchmark · ChestX-ray14, the original eight classes | AUC: Cardiomegaly 0.8141 -> 0.8100; Effusion 0.7362 -> 0.7585; Mass 0.5609 -> 0.6933; Nodule 0.7164 -> 0.6687 | Mixed |
Closest papers by meaning
- Bootstrapping Chest CT Image Understanding by Distilling Knowledge from X-ray Expert Models (CVPR 2024 · similarity 0.87)
- Learning Low-Rank Feature for Thorax Disease Classification (NeurIPS 2024 · similarity 0.86)
- BenchX: A Unified Benchmark Framework for Medical Vision-Language Pretraining on Chest X-Rays (NeurIPS 2024 · similarity 0.86)
- Novel Class Discovery in Chest X-rays via Paired Images and Text (AAAI 2024 · similarity 0.86)
- CXReasonBench: A Benchmark for Evaluating Structured Diagnostic Reasoning in Chest X-rays (NeurIPS 2025 · similarity 0.85)
- RadZero: Similarity-Based Cross-Attention for Explainable Vision-Language Alignment in Chest X-ray with Zero-Shot Multi-Task Capability (NeurIPS 2025 · similarity 0.85)
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.