Research map / Image generation / Text-to-image diffusion
ChEX: Interactive Localization and Region Description in Chest X-rays
Philip Müller, Georgios Kaissis, Daniel Rueckert
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 bounding box targets in multitask training (ablation) | ChEX · MS-CXR sentence grounding | mIoU: 3.68 -> 47.52 | Improved |
| add bounding box targets in multitask training (ablation) | ChEX · MIMIC-CXR report generation | Mic-F1-14: 40.68 -> 52.32 | Improved |
| add post decoder in the sentence generator (ablation) | ChEX · MIMIC-CXR report generation | Mic-F1-14: 46.07 -> 52.32 | Improved |
| add sentence tokens to the training mix (ablation) | ChEX · MS-CXR sentence grounding | mIoU: 52.97 -> 47.52 | Mixed |
| train with pathology, anatomy and sentence tokens together instead of only sentence tokens (ablation) | ChEX · VinDr-CXR pathology detection | mAP: 0.06 -> 14.12 | Improved |
| add anatomy tokens to the training mix (ablation) | ChEX · MIMIC-CXR report generation | Mic-F1-14: 46.64 -> 52.32 | Improved |
| add pathology classification labels (ablation) | ChEX · VinDr-CXR pathology detection | mAP: 10.35 -> 14.12 | Improved |
| use precise regional hints in textual prompts (e.g. pneumonia in the left lung) (ablation) | ChEX · pathology localization (gIoU) | localization and prediction accuracy | Improved |
| refine regional hint to finer region (upper lung) (ablation) | ChEX · pathology localization | localization quality | No clear effect |
| negate the prompt (ablation) | ChEX · pathology localization | localization quality | Hurt |
| use textual region prompts instead of bounding box queries (ablation) | ChEX · Chest ImaGenome region explanation | Mic-F1-14: 53.34 -> 54.70 | Improved |
| skip decoder language model pretraining on MIMIC-CXR (ablation) | ChEX sentence generator · MS-CXR region explanation | Mic-F1-14: 49.97 -> 48.56 | Hurt |
Closest papers by meaning
- CXReasonBench: A Benchmark for Evaluating Structured Diagnostic Reasoning in Chest X-rays (NeurIPS 2025 · similarity 0.90)
- BenchX: A Unified Benchmark Framework for Medical Vision-Language Pretraining on Chest X-Rays (NeurIPS 2024 · similarity 0.88)
- From Complexity to Clarity: Transforming Chest X-ray Reports with Chained Prompting (Student Abstract) (AAAI 2025 · similarity 0.87)
- AOR: Anatomical Ontology-Guided Reasoning for Medical Large Multimodal Model in Chest X-Ray Interpretation (NeurIPS 2025 · similarity 0.87)
- Rethinking Visual Attribution for Chest X-ray Reasoning in Large Vision Language Models (NeurIPS 2026 · similarity 0.87)
- MedRAX: Medical Reasoning Agent for Chest X-ray (ICML 2025 · similarity 0.87)
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.
