Research map / Vision-language and multimodal / Human and embodied modeling
KiUT: Knowledge-Injected U-Transformer for Radiology Report Generation
Read the paperCheck 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 |
|---|---|---|---|
| Use KiUT instead of AlignTrans | radiology report generation · MIMIC-CXR test | BLEU-1: 0.378 -> 0.393 | Helped |
| Use KiUT instead of XPRONET | radiology report generation · IU-Xray test | BLEU-4: 0.199 -> 0.185 | Hurt |
| Add intrinsic region relationships (ablation) | KiUT RR-Encoder on MIMIC-CXR · MIMIC-CXR | BLEU-1: 0.373 -> 0.382 | Helped |
| Add extrinsic region relationships (ablation) | KiUT RR-Encoder on MIMIC-CXR · MIMIC-CXR | BLEU-1: 0.373 -> 0.379 | Helped |
| Add contextual and clinical knowledge (ablation) | KiUT knowledge distiller · MIMIC-CXR | BLEU-1: 0.337 -> 0.393 | Helped |
| Add contextual knowledge (ablation) | KiUT knowledge distiller with clinical knowledge · MIMIC-CXR | BLEU-1: 0.371 -> 0.393 | Helped |
| Add clinical knowledge (ablation) | KiUT knowledge distiller with contextual knowledge · MIMIC-CXR | BLEU-1: 0.361 -> 0.393 | Helped |
| Replace last-layer-only connection with 1-to-1 encoder-decoder connections (ablation) | KiUT · MIMIC-CXR | BLEU-1: 0.372 -> 0.380 | Helped |
| Replace last-layer-only connection with U-connection (ablation) | KiUT · MIMIC-CXR | BLEU-1: 0.372 -> 0.393 | Helped |
| Replace last-layer-only connection with U-connection (ablation) | R2GenCMN · MIMIC-CXR | BLEU-1: 0.353 -> 0.363 | Helped |
| Replace last-layer-only connection with U-connection (ablation) | RSTNet · COCO test | BLEU-1: 0.811 -> 0.814 | Helped |
| Replace last-layer-only connection with U-connection (ablation) | M² Transformer · COCO test | BLEU-1: 0.808 -> 0.808 | No clear effect |
Closest papers by meaning
- AHIVE: Anatomy-aware Hierarchical Vision Encoding for Interactive Radiology Report Retrieval (CVPR 2024 · similarity 0.88)
- Instance-level Expert Knowledge and Aggregate Discriminative Attention for Radiology Report Generation (CVPR 2024 · similarity 0.88)
- Bootstrapping Large Language Models for Radiology Report Generation (AAAI 2024 · similarity 0.87)
- LLM-RG4: Flexible and Factual Radiology Report Generation Across Diverse Input Contexts (AAAI 2025 · similarity 0.86)
- Image-aware Evaluation of Generated Medical Reports (NeurIPS 2024 · similarity 0.85)
- DART: Disease-aware Image-Text Alignment and Self-correcting Re-alignment for Trustworthy Radiology Report Generation (CVPR 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.