Patch Ranking Map: Explaining Relations among Patches, Top-ranked Features and Decisions of a Convolutional Neural Network

Since a conventional Convolutional Neural Network (CNN) using a large number of extracted features is not fully explainable and not very memory-efficient, we develop a more explainable and efficient CNN model consisting of convolutional layers, a new feature selection (FS) layer, a classifier, and a novel "Patch Ranking Map" (PRM). The PRM contains top-ranked image patches that have important associations with decisions of the CNN. Top-ranked common features selected by different FS methods are used to generate two newly defined matrices: the "feature accumulation matrix" and the "feature ranking matrix". Different from a typical heatmap, these two matrices are used to rank image patches in the PRM to effectively explain the relationship among an input image, top-ranked features, top-ranked feature maps, and the final classification decision. Simulation results using the Alzheimer’s disease (AD) MRI preprocessed dataset for 4-class image classification with 6, 400 128 × 128 images indicate that the three PRMs based on three robust top-ranked common feature sets generated by seven different FS methods have the same top two most important patches associated with AD diagnosis. The PRM has explainable top-ranked patches because the top seven patches are associated with six brain areas associated with AD and three brain areas that may be associated with AD. The relationship among brain areas associated with AD, top-ranked patches, top patch rows, and top patch columns will be analyzed based on research results in artificial intelligence and medical informatics. The simulations also show that the trained CNN with FS can have higher classification accuracy and smaller model size than the conventional CNN without FS. More effective and efficient optimization algorithms will be developed to select the most informative features and rank patches for building an accurate and efficient CNN model with more explainable decisions that can be captured by the PRM for image classification applications.