Evidential Reasoning Advances Interpretable Real-World Disease Screening

Disease screening is critical for early detection and timely intervention in clinical practice. However, most current screening models for medical images suffer from limited interpretability and suboptimal performance. They often lack effective mechanisms to reference historical cases or provide transparent reasoning pathways. To address these challenges, we introduce EviScreen, an evidential reasoning framework for disease screening that leverages region-level evidence from historical cases. The proposed EviScreen offers retrospection interpretability through regional evidence retrieved from dual knowledge banks. Using this evidential mechanism, the subsequent evidence-aware reasoning module makes predictions using both the current case and evidence from historical cases, thereby enhancing disease screening performance. Furthermore, rather than relying on post-hoc saliency maps, EviScreen enhances localization interpretability by leveraging abnormality maps derived from contrastive retrieval. Our method achieves superior performance on our carefully established benchmarks for real-world disease screening, yielding notably higher specificity at clinical-level recall. Code is publicly available at https://github.com/DopamineLcy/EviScreen.

Catching diseases early is crucial, but current AI tools for analyzing medical images often act like "black boxes" that cannot explain their decisions. Furthermore, these systems rarely reference past patient cases to guide their current diagnoses. To solve this, we developed a new AI system that screens for diseases by mimicking how human clinicians work. When evaluating a new patient's scan, our tool searches a database of historical medical images to find similar visual evidence. It uses these past examples not only to make a more accurate prediction but also to clearly highlight exactly where and why it suspects an abnormality. This approach makes the AI's reasoning transparent and much easier for medical professionals to understand and trust. In comprehensive testing, our system shows superior performance, successfully reducing false alarms upon high recall scores for the patients who actually need care.