Accepted for/Published in: Journal of Medical Internet Research
Date Submitted: May 28, 2024
Date Accepted: Nov 29, 2024
Warning: This is an author submission that is not peer-reviewed or edited.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.
Detection of Alzheimer's Disease in Neuroimages using Vision Transformers: A Systematic Review and Meta-Analysis
ABSTRACT
Background:
Alzheimer's disease (AD) is a progressive neurodegenerative disorder, and early diagnosis is crucial for effective management and treatment. Vision Transformers (ViT) represent a cutting-edge advancement in deep learning, with significant potential for neuroimaging-based diagnostic applications.
Objective:
This systematic review and meta-analysis aim to evaluate the diagnostic accuracy of ViT-based models in detecting AD from neuroimaging data. We focus on different deep-learning network architectures and their comparative performance.
Methods:
A comprehensive search was conducted across major medical databases, including CNKI, CENTRAL, and ScienceDirect, up to March 1, 2024. This was supplemented by manual searches. The inclusion criteria encompassed studies using ViT models for AD detection versus healthy controls based on neuroimaging data. Pooled diagnostic accuracy estimates, including sensitivity, specificity, likelihood ratios, and diagnostic odds ratios, were derived using random-effects models. Subgroup analyses were performed to compare the diagnostic performance of different ViT network architectures.
Results:
A total of 11 studies were included in the meta-analysis. The pooled diagnostic accuracy was found to be high, with a sensitivity of 0.940 (95% CI: 0.903–0.963) and a specificity of 0.962 (95% CI: 0.933–0.978). The positive likelihood ratio was 26.02 (95% CI: 12.78–53.01), and the negative likelihood ratio was 0.07 (95% CI: 0.04–0.11). The area under the curve (AUC) was 0.9874, indicating excellent diagnostic performance.
Conclusions:
Our findings emphasize the significant potential of ViT-based models as effective tools for the early and accurate diagnosis of Alzheimer's disease. This systematic review provides robust evidence supporting the utility of ViTs in distinguishing AD patients from healthy controls, highlighting their promise in advancing neuroimaging-based diagnostic methodologies. Future research should focus on refining these models and addressing any barriers to clinical implementation. Clinical Trial: Not applicable
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.