Clinical knowledge guides clearer visual and text alignment for radiology reports
CliQ-RRG: Clinical-Knowledge Guided Disease-aware Visual-Textual Alignment for QA-Style Radiology Report Generation
Computation and Language
Summary
Writing radiology reports by hand takes a lot of time and mental effort. The authors created a two-step computer method called CliQ-RRG that helps produce these reports automatically. First, the method connects X-ray images with disease information to understand what’s important across different scans. Then, it retrieves related medical facts and turns this into easy-to-understand question and answer pairs summarizing the diagnosis. They tested this approach on chest X-ray data and found it made more accurate and clear reports than earlier methods.
What this means in practice
- •For radiology software developers: Integrate disease-aware image-text alignment and clinical knowledge retrieval to create more accurate, structured radiology reports automatically.
- •For medical imaging platform teams: Add question-answer style summaries with multi-view scan context to help doctors interpret radiology reports more efficiently.
Authors
Chandravardhan Singh Raghaw, Shahid Shafi Dar, Arnav Nishith Sharma, Shrey Modi, Nagendra Kumar
Abstract
Radiology reports are essential for diagnostic reasoning and patient care, yet their manual preparation is time-consuming and cognitively demanding. Automatic radiology report generation (RRG) offers a scalable alternative, but existing models often produce lengthy, unstructured narratives that overlook diagnostic cues and multi-view information. We present Clinical-Knowledge Guided Disease-aware Visual-Textual Alignment for QA-Style Radiology Report Generation (CliQ-RRG), a unified two-stage framework for interpretable and clinically grounded reporting. In Stage 1, CliQ-RRG employs a Disease-aware Visual-Textual Alignment module that aligns image and text representations using predicted disease embeddings, reinforced by a Prior-Guided Attention Module (PrAM) to capture multi-view dependencies across current and prior scans. In Stage 2, GraphRAG-guided retrieval augments intermediate textual representations with graph-grounded clinical evidence, and a large language model restructures them into concise, interpretable question-answer (QA) pairs with diagnostic summaries. Experiments on two public chest X-ray benchmarks demonstrate that CliQ-RRG consistently outperforms prior methods across both lexical and clinical metrics, generating accurate and clinically coherent QA-style radiology reports. Code is available at: https://github.com/chandravardhan-singh-raghaw/CliQ-RRG.