Towards Trainable Saliency Maps in Medical Imaging
This addresses the need for interpretable AI in medical diagnosis, particularly for improving acceptability and safety, though it appears incremental as it builds on existing saliency methods.
The paper tackles the problem of black-box decision-making in deep learning for medical imaging by introducing a model design element that creates inherently self-explanatory models, demonstrating higher localization efficacy on the RSNA Pneumonia Dataset compared to non-trainable saliency maps and offering a reasonable alternative to fully supervised baselines with high data labeling overhead.
While success of Deep Learning (DL) in automated diagnosis can be transformative to the medicinal practice especially for people with little or no access to doctors, its widespread acceptability is severely limited by inherent black-box decision making and unsafe failure modes. While saliency methods attempt to tackle this problem in non-medical contexts, their apriori explanations do not transfer well to medical usecases. With this study we validate a model design element agnostic to both architecture complexity and model task, and show how introducing this element gives an inherently self-explanatory model. We compare our results with state of the art non-trainable saliency maps on RSNA Pneumonia Dataset and demonstrate a much higher localization efficacy using our adopted technique. We also compare, with a fully supervised baseline and provide a reasonable alternative to it's high data labelling overhead. We further investigate the validity of our claims through qualitative evaluation from an expert reader.