Abstract
The timely identification of Alzheimer’s Disease (AD) is crucial for the provision of adequate treatment to the patients. The traditional diagnostic techniques such as neuroimaging, CSF examination, and neuropsychological testing might be relatively expensive, invasive, and sensitive. Due to the development of Artificial Intelligence (AI) and Deep Learning technologies, it is now possible to develop more effective and precise approaches to AD diagnostics.The goal of the paper is to present a review of the latest researches on the application of AI for the purposes of AD diagnostics, paying special attention to multimodal Vision Transformers (ViT) and its application. Multimodal ViTs demonstrate high efficiency when it comes to recognition of complex visual and linguistic features. In addition to that, multimodal fusion allows combining different types of information received from such sources as MRI and PET scans, language, handwriting, and biomarker tests. The findings of the most relevant papers are analyzed in detail and compared in terms of performance and computing resources required.In general, multimodal ViTs provide better results than common single-modal models and CNNs in diagnosing AD according to the findings of several researches.