🤖 AI Summary
Traditional end-to-end sequence modeling for medical report generation suffers from limited diagnostic consistency and poor interpretability. This paper reformulates the task as a knowledge graph–guided multi-label classification problem, leveraging a radiology-specific knowledge graph to identify key imaging findings (e.g., “pulmonary consolidation”, “pleural effusion”)—enabling structured, spatially localizable pathological representation. Methodologically, we extend the BLIP architecture with a knowledge graph–guided multi-label classifier and a node-conditioned report reconstruction mechanism, jointly optimizing generation fidelity and clinical interpretability. Evaluated on the IU-Xray and MIMIC-CXR benchmarks, our approach achieves state-of-the-art performance, improving diagnostic consistency by +4.2% and report accuracy by +3.8%, while reducing inference complexity.
📝 Abstract
Medical report generation is a critical task in healthcare that involves the automatic creation of detailed and accurate descriptions from medical images. Traditionally, this task has been approached as a sequence generation problem, relying on vision-and-language techniques to generate coherent and contextually relevant reports. However, in this paper, we propose a novel perspective: rethinking medical report generation as a multi-label classification problem. By framing the task this way, we leverage the radiology nodes from the commonly used knowledge graph, which can be better captured through classification techniques. To verify our argument, we introduce a novel report generation framework based on BLIP integrated with classified key nodes, which allows for effective report generation with accurate classification of multiple key aspects within the medical images. This approach not only simplifies the report generation process but also significantly enhances performance metrics. Our extensive experiments demonstrate that leveraging key nodes can achieve state-of-the-art (SOTA) performance, surpassing existing approaches across two benchmark datasets. The results underscore the potential of re-envisioning traditional tasks with innovative methodologies, paving the way for more efficient and accurate medical report generation.