Interpretable Fine-Gray Deep Survival Model for Competing Risks: Predicting Post-Discharge Foot Complications for Diabetic Patients in Ontario
This work addresses the need for transparent AI in clinical practice for diabetic patients, though it is incremental as it builds on existing neural additive models.
The authors tackled the problem of limited interpretability in deep survival models for competing risks by proposing CRISPNAM-FG, an intrinsically interpretable model that predicts foot complications in diabetic patients with competitive performance compared to other deep methods.
Model interpretability is crucial for establishing AI safety and clinician trust in medical applications for example, in survival modelling with competing risks. Recent deep learning models have attained very good predictive performance but their limited transparency, being black-box models, hinders their integration into clinical practice. To address this gap, we propose an intrinsically interpretable survival model called CRISPNAM-FG. Leveraging the structure of Neural Additive Models (NAMs) with separate projection vectors for each risk, our approach predicts the Cumulative Incidence Function using the Fine-Gray formulation, achieving high predictive power with intrinsically transparent and auditable predictions. We validated the model on several benchmark datasets and applied our model to predict future foot complications in diabetic patients across 29 Ontario hospitals (2016-2023). Our method achieves competitive performance compared to other deep survival models while providing transparency through shape functions and feature importance plots.