Evidential Reasoning Advances Interpretable Real-World Disease Screening

TL;DR

EviScreen is an evidential reasoning framework that improves interpretability and performance in disease screening by leveraging historical case evidence and abnormality maps.

Contribution

The paper introduces EviScreen, a novel framework that combines regional evidence retrieval and contrastive abnormality maps for interpretable and accurate disease screening.

Findings

EviScreen achieves higher specificity at clinical recall levels.

The framework enhances localization interpretability without post-hoc saliency maps.

Superior performance demonstrated on real-world disease screening benchmarks.

Abstract

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.