Layer-Wise Relevance Propagation for Explaining Deep Neural Network Decisions in MRI-Based Alzheimer's Disease Classification
Послойное распространение значимости для объяснения решений глубоких нейронных сетей при классификации болезни Альцгеймера по данным МРТ
2019-07-31
SCID: 54.1/u32a8j98
Discuss with AI
Alzheimer’s disease classificationConvolutional neural networksLayer-wise relevance propagationRelevance densityStructural MRI
Figures from the paper
Abstract (AI)
Deep neural networks have led to state-of-the-art results in many medical imaging tasks including Alzheimer's disease (AD) detection based on structural magnetic resonance imaging (MRI) data. However, the network decisions are often perceived as being highly non-transparent, making it difficult to apply these algorithms in clinical routine. In this study, we propose using layer-wise relevance propagation (LRP) to visualize convolutional neural network decisions for AD based on MRI data. Similarly to other visualization methods, LRP produces a heatmap in the input space indicating the importance/relevance of each voxel contributing to the final classification outcome. In contrast to susceptibility maps produced by guided backpropagation ("Which change in voxels would change the outcome most?"), the LRP method is able to directly highlight positive contributions to the network classification in the input space. In particular, we show that (1) the LRP method is very specific for individuals ("Why does this person have AD?") with high inter-patient variability, (2) there is very little relevance for AD in healthy controls and (3) areas that exhibit a lot of relevance correlate well with what is known from literature. To quantify the latter, we compute size-corrected metrics of the summed relevance per brain area, e.g., relevance density or relevance gain. Although these metrics produce very individual "fingerprints" of relevance patterns for AD patients, a lot of importance is put on areas in the temporal lobe including the hippocampus. After discussing several limitations such as sensitivity toward the underlying model and computation parameters, we conclude that LRP might have a high potential to assist clinicians in explaining neural network decisions for diagnosing AD (and potentially other diseases) based on structural MRI data.
Key Findings
1
Brain regions with high relevance align with established Alzheimer’s disease literature, particularly temporal-lobe areas including the hippocampus.
2
LRP explanations are highly individual-specific, showing substantial inter-patient variability and very little Alzheimer’s-related relevance in healthy controls.
3
Layer-wise relevance propagation (LRP) visualizes convolutional neural network decisions for MRI-based Alzheimer’s disease classification through voxel-level relevance heatmaps.
4
Size-corrected relevance metrics, such as relevance density and relevance gain, quantify individual relevance patterns but remain sensitive to the underlying model and computation parameters.
5
Unlike guided backpropagation susceptibility maps, LRP directly identifies voxels making positive contributions to an Alzheimer’s disease classification.
Research Object
MRI-based deep neural network classification of Alzheimer's disease
Research Subject
Layer-wise relevance patterns and voxel-level contributions underlying individual neural network decisions for Alzheimer's disease, including regional relevance density and relevance gain
Publication Details
Publication Date
2019-07-31
Journal
Publisher
ISSN
Cited by
291
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai4
Cited by8
Transparency of deep neural networks for medical image analysis: A review of interpretability methods2021
Explainable, trustworthy, and ethical machine learning for healthcare: A survey2022
Survey of explainable artificial intelligence techniques for biomedical imaging with deep neural networks2023
Explainable Artificial Intelligence in Alzheimer’s Disease Classification: A Systematic Review2023
Explainable AI for Bioinformatics: Methods, Tools and Applications2023
Explainable AI: A review of applications to neuroimaging data2022
Explainable Artificial Intelligence in Neuroimaging of Alzheimer’s Disease2025
Explainable Deep CNNs for MRI-Based Diagnosis of Alzheimer’s Disease2020