communication mediumInternational Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2020
The success of machine learning methods for computer vision tasks has driven a surge in computer assisted prediction for medicine and biology. Based on a data-driven relationship between input image and pathological classiﬁcation, these predictors deliver unprecedented accuracy. Yet, the numerous approaches trying to explain the causality of this learned relationship have fallen short: time constraints, coarse, diﬀuse and at times misleading results, caused by the employment of heuristic techniques like Gaussian noise and blurring, have hindered their clinical adoption. In this work, we discuss and overcome these obstacles by introducing a neural-network based attribution method, applicable to any trained predictor. Our solution identiﬁes salient regions of an input image in a single forward-pass by measuring the eﬀect of local image-perturbations on a predictor’s score. We replace heuristic techniques with a strong neighborhood conditioned inpainting approach, avoiding anatomically implausible, hence adversarial artifacts. We evaluate on public mammography data and compare against existing state-of-the-art methods. Furthermore, we exemplify the approach’s generalizability by demonstrating results on chest X-rays. Our solution shows, both quantitatively and qualitatively, a signiﬁcant reduction of localization ambiguity and clearer conveying results, without sacriﬁcing time eﬃciency.
Explainable AI; XAI; Classifier Decision Visualization; Image Impainting