Diagrammatization and Abduction to Improve AI Interpretability With Domain-Aligned Explanations for Medical Diagnosis
This work addresses the interpretability gap in high-stakes medical diagnosis for expert users, offering domain-aligned explanations, though it is incremental as it builds on existing XAI methods.
The paper tackled the problem of improving interpretability in AI for medical diagnosis by introducing DiagramNet, which uses diagrammatic and abductive reasoning to provide clinically-relevant explanations for cardiac diagnoses from heart auscultation, resulting in better performance than baseline models and a preference for diagrammatic explanations over saliency maps in a user study.
Many visualizations have been developed for explainable AI (XAI), but they often require further reasoning by users to interpret. Investigating XAI for high-stakes medical diagnosis, we propose improving domain alignment with diagrammatic and abductive reasoning to reduce the interpretability gap. We developed DiagramNet to predict cardiac diagnoses from heart auscultation, select the best-fitting hypothesis based on criteria evaluation, and explain with clinically-relevant murmur diagrams. The ante-hoc interpretable model leverages domain-relevant ontology, representation, and reasoning process to increase trust in expert users. In modeling studies, we found that DiagramNet not only provides faithful murmur shape explanations, but also has better performance than baseline models. We demonstrate the interpretability and trustworthiness of diagrammatic, abductive explanations in a qualitative user study with medical students, showing that clinically-relevant, diagrammatic explanations are preferred over technical saliency map explanations. This work contributes insights into providing domain-aligned explanations for user-centric XAI in complex domains.