Alzheimer's disease detection using generative adversarial networks: a survey

Alzheimer's disease (AD) is a chronic neurodegenerative disorder that progressively affects memory, cognition, and overall functional ability. Timely and accurate diagnosis is crucial for effective management and slowing the disease's progression. Traditional diagnostic methods – such as cognitive evaluations, neuroimaging techniques (e.g. MRI and PET), and biomarker testing often detect AD only after the onset of clinical symptoms, limiting early intervention opportunities. In recent years, advances in machine learning have demonstrated significant potential in identifying Alzheimer's-related patterns by analyzing large and complex biomedical datasets, including neuroimaging, genetic, and biochemical data. Deep learning, a subset of machine learning, offers enhanced capabilities in processing such data, with Generative Adversarial Networks (GANs) emerging as a particularly promising approach. GANs are increasingly being explored for their ability to synthesize and augment multimodal data, enhance feature extraction, and to improve classification performance in AD detection tasks. This survey provides a comprehensive review of current research leveraging GANs for Alzheimer's disease detection. We examine various methodological frameworks, summarize key findings, highlight existing limitations, and discuss open challenges in the field. Finally, we propose future research directions aimed at advancing the role of GANs in early and accurate diagnosis of Alzheimer's disease.