Accepted for/Published in: JMIR Medical Informatics
Date Submitted: Feb 4, 2025
Open Peer Review Period: Feb 10, 2025 - Apr 7, 2025
Date Accepted: Jul 30, 2025
(closed for review but you can still tweet)
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Interpretable Analysis of a Functional Magnetic Resonance Imaging-Based Migraine Classification Model
ABSTRACT
Background:
Deep learning has shown great promise in advancing computer-aided diagnosis for neuropsychiatric diseases like migraines. However, the ability to interpret the image classification results lags far behind, confounding their clinical translation.
Objective:
To utilize explainable artificial intelligence (XAI) methods combined with different MRI indicators to generate brain activation heatmaps, compare the effectiveness of different fMRI indicators and network architectures in classifying migraines, determine the best combination to enhance classification performance, and assess the potential of XAI technology in clinical diagnosis.
Methods:
The study included scans from 21 migraine patients without aura, 15 patients with aura, and 28 healthy controls. Various fMRI indicators [amplitude of low-frequency fluctuations (ALFF), regional homogeneity (ReHo), and regional functional correlation strength (RFCS)] and three network architectures (GoogleNet, ResNet18, and ViT-B/16) were used to classify migraines. The classification performance of each model was compared, and heatmaps were generated using Grad-CAM or self-attention mechanisms to identify discriminative regions.
Results:
The GoogleNet model combined with RFCS indicators achieved the best classification performance, with an accuracy of over 98.44% and an area under the receiver operating characteristic curve (AUC) of 0.99 for the test set. Additionally, among the three indicators, the RFCS indicator improved accuracy by about 8% compared to ALFF. Brain activation heatmaps generated by XAI technology revealed that the precuneus and cuneus were the most discriminative brain regions, with slight activation also observed in the frontal gyrus.
Conclusions:
The use of XAI technology combined with brain region features provides visual explanations for the progression of migraine in patients. Understanding the decision-making process of the network has significant potential for clinical diagnosis of migraines, offering promising applications in enhancing diagnostic accuracy and aiding in the development of new diagnostic techniques.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.