Explainable Deep CNNs for MRI-Based Diagnosis of Alzheimer’s Disease
Интерпретируемые глубокие сверточные нейронные сети для диагностики болезни Альцгеймера на основе МРТ
2020-07-01
SCID: 54.1/d9uwm57n
Discuss with AI
Alzheimer’s disease diagnosisSwap Testaxiomatic evaluationbrain MRIexplainable deep CNNs
Figures from the paper
Abstract (AI)
Deep Convolutional Neural Networks (CNNs) are becoming prominent models for semi-automated diagnosis of Alzheimer's Disease (AD) using brain Magnetic Resonance Imaging (MRI). Although being highly accurate, deep CNN models lack transparency and interpretability, precluding adequate clinical reasoning and not complying with most current regulatory demands. One popular choice for explaining deep image models is occluding regions of the image to isolate their influence on the prediction. However, existing methods for occluding patches of brain scans generate images outside the distribution to which the model was trained for, thus leading to unreliable explanations. In this paper, we propose an alternative explanation method that is specifically designed for the brain scan task. Our method, which we refer to as Swap Test, produces heatmaps that depict the areas of the brain that are most indicative of AD, providing interpretability for the model's decisions in a format understandable to clinicians. Experimental results using an axiomatic evaluation show that the proposed method is more suitable for explaining the diagnosis of AD using MRI while the opposite trend was observed when using a typical occlusion test. Therefore, we believe our method may address the inherent black-box nature of deep neural networks that are capable of diagnosing AD.
Key Findings
1
Axiomatic evaluation found Swap Test more suitable than typical occlusion testing for explaining Alzheimer’s disease diagnosis from MRI.
2
Existing patch-occlusion explanations for brain MRI CNNs can generate out-of-distribution images, making their diagnostic explanations unreliable.
3
Swap Test explanations are presented in a clinician-understandable format, improving interpretability of CNN-based MRI diagnoses.
4
The method aims to reduce the black-box nature of deep CNNs while supporting clinically interpretable Alzheimer’s disease diagnosis.
5
The proposed Swap Test is specifically designed for brain scans and generates heatmaps highlighting brain regions most indicative of Alzheimer’s disease.
Research Object
brain MRI-based deep CNN diagnosis of Alzheimer’s disease
Research Subject
interpretability of CNN diagnostic decisions, specifically identifying brain regions indicative of Alzheimer’s disease through reliable heatmap explanations
Publication Details
Publication Date
2020-07-01
Journal
Publisher
ISSN
Cited by
56
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai7
Deep Residual Learning for Image Recognition2016
ImageNet classification with deep convolutional neural networks2017
Very Deep Convolutional Networks for Large-Scale Image Recognition2014
Adam: A Method for Stochastic Optimization2014
Gradient-based learning applied to document recognition1998
Learning Deep Features for Discriminative Localization2016
Layer-Wise Relevance Propagation for Explaining Deep Neural Network Decisions in MRI-Based Alzheimer's Disease Classification2019