Advances in structural magnetic resonance imaging for the diagnosis of Alzheimer's disease

Alzheimer's disease (AD) is a progressive neurodegenerative disorder mainly characterised by developing cognitive damage and non-cognitive decline. Structural magnetic resonance imaging (sMRI) is an important diagnostic tool for clinical diagnosis by detecting structural changes in the brain to reveal early pathological features of AD. However, there are several challenges in detecting Alzheimer's disease, including structural variability, mixed pathology, early lesion detection, and long-term monitoring, but there are insufficient effective approaches to achieve them, all of which require advanced technological approaches for accurate diagnosis. Various deep learning techniques have been suggested for automated classification of AD from 3D brain sMRI. These methods are able to automatically identify characteristic changes in AD by learning from a large amount of image data of the AD brain. Nevertheless, articles combining anatomical pathology and deep learning are still very rare. This paper reviews the research progress of structural MRI in the context of AD’s diagnosis and discusses the future prospects of AD diagnosis. By summarising the diagnostic pathways related to pathology, combined with current deep learning methods, we hope that the article will provide further insights and guidance within the domain of medicine and deep learning, leading to new breakthroughs in the identification and treatment of AD.