Exploring Regions of Interest: Visualizing Histological Image Classification for Breast Cancer using Deep Learning
This work addresses the need for interpretability in computer-aided diagnosis systems for breast cancer, though it is incremental as it applies existing visualization methods to a specific domain.
The study tackled the problem of justifying deep learning classifications in breast cancer detection by visualizing regions of interest used by a CNN for histological image classification, comparing them with pathologist-identified regions and finding that Gradient visualization with MeanShift selection yielded satisfactory outcomes.
Computer aided detection and diagnosis systems based on deep learning have shown promising performance in breast cancer detection. However, there are cases where the obtained results lack justification. In this study, our objective is to highlight the regions of interest used by a convolutional neural network (CNN) for classifying histological images as benign or malignant. We compare these regions with the regions identified by pathologists. To achieve this, we employed the VGG19 architecture and tested three visualization methods: Gradient, LRP Z, and LRP Epsilon. Additionally, we experimented with three pixel selection methods: Bins, K-means, and MeanShift. Based on the results obtained, the Gradient visualization method and the MeanShift selection method yielded satisfactory outcomes for visualizing the images.