Although deep learning networks applied to digital images have shown impressive results for many pathology-related tasks, their black-box approach and limitation in terms of interpretability are significant obstacles for their widespread clinical utility. This study investigates the visualization of deep features to characterize two lung cancer subtypes, adenocarcinoma, and squamous cell carcinoma. This study demonstrates that a subset of deep features exist that can accurately distinguish these two cancer subtypes, “prominent deep features.” Visualization of such individual deep features allows us to understand better histopathologic patterns at both the whole-slide and patch levels allowing discrimination of these cancer types. These deep features were visualized at the whole slide image-level through deep feature-specific heatmaps and at tissue patch level through generating activation maps. Additionally, we show that these prominent deep features contain information that can distinguish carcinomas of organs other than the lung. This framework may serve as a platform for evaluating the interpretability of any deep network for diagnostic decision-making.
Copyright © 2021. Published by Elsevier Inc.

Author