Advances in multimodal deep learning for early detection of Alzheimer's disease
10.3969/j.issn.1005-202X.2025.01.004
- VernacularTitle:阿尔茨海默病早期检测的多模态深度学习研究进展
- Author:
DI LI
1
;
Xufeng YAO
Author Information
1. 上海理工大学健康科学与工程学院,上海200093;上海健康医学院医学影像学院,上海201318
- Publication Type:Journal Article
- Keywords:
Alzheimer's disease;
multimodal;
deep learning;
neuroimaging
- From:
Chinese Journal of Medical Physics
2025;42(1):20-26
- CountryChina
- Language:Chinese
-
Abstract:
Alzheimer's disease (AD) is a chronic neurodegenerative disease that mainly affects neurons in the brain,especially in regions related to memory,thinking,and behavior. During the auxiliary diagnosis of AD,massive data from imaging,genetics,transcriptomics as well as clinical features provide a new basis for mining potential molecular markers and the early diagnosis and intervention of AD. In recent years,deep learning models have shown strong feature learning and prediction capabilities in AD image classification;and the researchers will effectively integrate various modal data to provide richer complementary information for further improving the classification performance. Herein the review introduces the commonly used neuroimaging data sets and evaluation criteria for AD,analyzes the application of various modal data in AD classification,focusing on the application of multimodal data in AD classification diagnosis,discuss the application of the classic deep learning network model in AD classification diagnosis,aiming to provide ideas for further research on multimodal deep learning technology.
