2019/09/30 by Irina Grigorescu, Grigorescu, Irina, Lucilio Cordero‐Grande +10 · 1 citation
Computer Science · Medicine · #Advanced Neural Network Applications #Domain Adaptation and Few-Shot Learning #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Machine Learning in Healthcare #Neonatal and fetal brain pathology #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1910.00071
openalex publication_date 2019/09/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The use of convolutional neural networks (CNNs) for classification tasks has\nbecome dominant in various medical imaging applications. At the same time,\nrecent advances in interpretable machine learning techniques have shown great\npotential in explaining classifiers' decisions. Layer-wise relevance\npropagation (LRP) has been introduced as one of these novel methods that aim to\nprovide visual interpretation for the network's decisions. In this work we\npropose the application of 3D CNNs with LRP for the first time for neonatal\nT2-weighted magnetic resonance imaging (MRI) data analysis. Through LRP, the\ndecisions of our trained classifier are transformed into heatmaps indicating\neach voxel's relevance for the outcome of the decision. Our resulting LRP\nheatmaps reveal anatomically plausible features in distinguishing preterm\nneonates from term ones.\n