Research map / Vision-language and multimodal / Human and embodied modeling
Interactive and Explainable Region-Guided Radiology Report Generation
Tim Tanida, Philip Müller, Georgios Kaissis, Daniel Rueckert
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 |
|---|---|---|---|
| Use region-guided generation with anatomy detection, selection, and region-specific sentences | RGRG · MIMIC-CXR | METEOR: 0.158 -> 0.168 | Improved |
| Use region-guided generation with anatomy detection, selection, and region-specific sentences | RGRG · MIMIC-CXR | ROUGE-L: 0.285 -> 0.264 | Hurt |
| Use region-guided generation with anatomy detection, selection, and region-specific sentences | RGRG · MIMIC-CXR | F1, ex-14: 0.384 -> 0.447 | Improved |
| Add region selection to the detector-plus-language-model base (ablation) | RGRG · MIMIC-CXR | BLEU-4: 0.104 -> 0.114 | Improved |
| Add region selection to the detector-plus-language-model base (ablation) | RGRG · MIMIC-CXR | F1, mic-5: 0.443 -> 0.523 | Improved |
| Add abnormality classification to the detector-plus-language-model base (ablation) | RGRG · MIMIC-CXR | F1, mic-5: 0.443 -> 0.501 | Improved |
| Add abnormality classification to the detector-plus-language-model base (ablation) | RGRG · MIMIC-CXR | BLEU-4: 0.104 -> 0.107 | No clear effect |
| Add abnormality classification to the detector-plus-language-model base (ablation) | RGRG · MIMIC-CXR | R, mic-5: 0.359 -> 0.461 | Improved |
| Add both region selection and abnormality classification (ablation) | RGRG · MIMIC-CXR | F1, mic-5: 0.443 -> 0.547 | Improved |
| Use RoI feature vectors instead of pooled RoI feature maps and a linear transformation (ablation) | RGRG object detector · MIMIC-CXR | object detector performance | Hurt |
| Vary the manually drawn box position away from the target region (ablation) | RGRG selection-based sentence generation · Chest ImaGenome | METEOR | Hurt |
| Vary manually drawn box aspect ratio (ablation) | RGRG selection-based sentence generation · Chest ImaGenome | METEOR | No clear effect |
Closest papers by meaning
- LLM-RG4: Flexible and Factual Radiology Report Generation Across Diverse Input Contexts (AAAI 2025 · similarity 0.88)
- Bootstrapping Large Language Models for Radiology Report Generation (AAAI 2024 · similarity 0.87)
- S2D-Align: Shallow-to-Deep Auxiliary Learning for Anatomically-Grounded Radiology Report Generation (AAAI 2026 · similarity 0.87)
- AHIVE: Anatomy-aware Hierarchical Vision Encoding for Interactive Radiology Report Retrieval (CVPR 2024 · similarity 0.87)
- Instance-level Expert Knowledge and Aggregate Discriminative Attention for Radiology Report Generation (CVPR 2024 · similarity 0.86)
- Reporting Practice Matters: The Impact of Reference Choice on Chest X-ray Report Evaluation (NeurIPS 2026 · similarity 0.86)
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.
