Search papers, labs, and topics across Lattice.
This paper introduces RadFusion, a novel framework for automated radiology report generation that allows for threshold-controllable sensitivity and specificity in diagnostic content. By integrating a multi-label classifier with a visual question answering (VQA) report generator and an LLM for report rewriting, RadFusion adapts report outputs to clinical scenarios while conforming to ROC curve performance. The results demonstrate significant improvements in diagnostic accuracy, with sensitivity increasing by 6.9% and specificity by 20.7% at matched thresholds, enhancing both clinical adaptability and regulatory validation potential.
RadFusion enables radiology reports to flexibly adapt their sensitivity and specificity based on clinical needs, revolutionizing automated diagnostic reporting.
Automated radiology report generation is advancing rapidly in response to the shortage of radiologists, yet unlike a perception model, existing generation models offer no control over the sensitivity-specificity trade-off of their diagnostic content. Such control is essential because clinical scenarios diverge: emergency triage prioritizes sensitivity to reduce missed findings, whereas confirmatory interpretation emphasizes specificity to limit unnecessary interventions. A single fixed report can neither adapt to these scenarios nor support the ROC-based validation widely expected for regulatory clearance. We introduce RadFusion, a framework that equips report generation with threshold controllability. Our method fuses a multi-label classifier, which provides per-disease confidence scores, with a VQA-based report generator, which describes medical findings in detail; an LLM then rewrites the report so that its stated diagnoses follow the classifier's decisions at the selected threshold while staying grounded in the generator's descriptions. On MIMIC-CXR, the performance of RadFusion conforms to the classifier's ROC curve: sweeping the threshold and mapping the reports back to class labels reproduces the classifier's validated ROC performance. This conformance makes generated reports quantitatively evaluable through ROC analysis, strengthening the case for regulatory clearance, and enables operating-point selection that matches report behavior to clinical context. Moreover, combining the two model types improves diagnostic accuracy over uncontrolled generation: sensitivity increases by 6.9% at matched specificity, and specificity by 20.7% at matched sensitivity. These results show that RadFusion makes report generation clinically adaptable, quantitatively verifiable, and diagnostically more reliable.
Microsoft Research