No Black Box Anymore: Demystifying Clinical Predictive Modeling with Temporal-Feature Cross Attention Mechanism
This addresses the 'black box' limitation in healthcare AI by providing transparent insights for clinicians, though it is an incremental improvement in explainability methods.
The paper tackled the problem of explainability in deep learning for clinical prediction by introducing the Temporal-Feature Cross Attention Mechanism (TFCAM), which achieved an AUROC of 0.95 and an F1-score of 0.69 in predicting End-Stage Renal Disease progression in 1,422 patients, outperforming baselines.
Despite the outstanding performance of deep learning models in clinical prediction tasks, explainability remains a significant challenge. Inspired by transformer architectures, we introduce the Temporal-Feature Cross Attention Mechanism (TFCAM), a novel deep learning framework designed to capture dynamic interactions among clinical features across time, enhancing both predictive accuracy and interpretability. In an experiment with 1,422 patients with Chronic Kidney Disease, predicting progression to End-Stage Renal Disease, TFCAM outperformed LSTM and RETAIN baselines, achieving an AUROC of 0.95 and an F1-score of 0.69. Beyond performance gains, TFCAM provides multi-level explainability by identifying critical temporal periods, ranking feature importance, and quantifying how features influence each other across time before affecting predictions. Our approach addresses the "black box" limitations of deep learning in healthcare, offering clinicians transparent insights into disease progression mechanisms while maintaining state-of-the-art predictive performance.