Beyond AT(N): Integrating Neuroimaging, Fluid Biomarkers, and Artificial Intelligence for Alzheimer’s Disease Diagnosis—A Review

Résumé fourni par la source

This review consists of three parts related to the AT(N) framework, providing a comprehensive overview of biomarker-based diagnosis of Alzheimer’s disease (AD). In recent years, management of AD has shifted from a symptom-based approach toward biological biomarkers targeted at the amyloid (A), tau (T), and neurodegeneration (N) biomarkers, collectively known as the AT(N) framework. This shift has led to the development and investigation of numerous biomarkers with diverse applications and orientations in the literature. The first part of this review focuses on the image-based biomarkers across all available AD neuroimaging modalities and discusses their roles within the AT(N) framework. The second part concentrates on the emerging fluid-based biomarkers derived from cerebrospinal fluid (CSF) and blood samples and how they are incorporated into the AT(N) framework, leading to the recommendation of the ATN(X) model, which is an expanded biomarker classification incorporating additional pathophysiological processes and is currently undergoing clinical validation. Together, both parts summarize the role and potential of every individual image-based or fluid-based biomarker in the diagnosis, classification, and management of AD within the AT(N) and ATN(X) frameworks. Then, the third part revises major recent developments in artificial intelligence (AI)-based approaches and explains how the various imaging- and fluid-based AD biomarkers are integrated into different AI models, learning strategies, and interpretability methods for the AD diagnosis. Collectively, this review establishes a comprehensive synthesis of the current state-of-the-art image-based and fluid-based AD biomarkers, highlighting their roles in biomarker-based AD diagnosis, the AT(N)/ATN(X) framework, and (AI)-based applications. Furthermore, this three-part review emphasizes that, despite the need for greater standardization, improved interpretability, increased data availability, and further clinical validation, integrated imaging- and fluid-based biomarker–AI approaches hold considerable promise for advancing AD diagnosis and management.