The National Institutes of Health's renewed support for the Artificial Intelligence for Alzheimer's Disease (AI4AD) initiative, with a substantial $12.6 million investment, marks a significant leap forward in the quest to unravel the complexities of Alzheimer's and related dementias. This funding, bringing the total to $30.7 million, is a testament to the project's potential to revolutionize our understanding and treatment of these devastating neurological conditions. Led by the esteemed Paul M. Thompson, PhD, the AI4AD2 project is a multi-institutional endeavor, uniting 10 investigators and 23 co-investigators from diverse institutions. The primary objective is to harness the power of artificial intelligence to delve into the biological underpinnings of Alzheimer's and related dementias, enhance disease progression predictions, and pave the way for more tailored treatment strategies.
One of the project's initial goals is to transcend broad diagnostic labels and delve into the identification of meaningful subtypes of Alzheimer's disease and related dementias. By employing AI to categorize individuals based on intricate patterns in brain scans, cognition, neuropathology, and genetic data, the project aims to improve clinical trial design. This molecular subtyping becomes increasingly crucial as new therapies target specific biological processes, such as amyloid, tau, vascular injury, and inflammation, which vary in their impact on different patients.
AI4AD2 introduces a novel concept in the form of 'genomic language models,' an AI approach inspired by language-based systems. These models will analyze genomic sequences to pinpoint DNA changes associated with Alzheimer's disease, progression, and key biomarkers. By training and evaluating these methods using extensive data from over 58,000 participants across 57 cohorts, the project aims to uncover novel genetic and protein-related changes driving neurodegeneration. This builds upon earlier research demonstrating AI's ability to identify Alzheimer's-related features on brain scans with over 90% accuracy, achieved through the analysis of 80,000 brain scans.
Another critical aspect of AI4AD2 is ensuring the global applicability of its AI tools. Recognizing the limitations of existing biomedical datasets focused on European ancestry, the project will adapt its disease classification, subtyping, and prognosis tools for diverse populations, including African, Indian, Korean, and US datasets. This approach aims to identify how ancestry, social, and environmental factors influence Alzheimer's risk and progression, ultimately leading to more accurate predictive models.
The project's fourth goal revolves around the discovery of treatments through genome-guided drug discovery. Utilizing the PreSiBO system, an AI-based drug discovery tool developed through the original AI4AD effort, researchers will identify subtype-specific therapeutic targets and evaluate the potential of repurposing existing drugs for patients with specific Alzheimer's-related biological profiles. This involves developing AI tools to detect multiple molecular pathways and pinpoint specific drug treatments targeting these pathways.
The USC Mark and Mary Stevens Neuroimaging and Informatics Institute (Stevens INI) will serve as a pivotal hub for this collaborative endeavor. AI4AD2 is designed to foster collaboration, with USC as the lead site and partner institutions contributing expertise in neuroimaging, genomics, statistics, machine learning, cognitive science, and drug discovery. The team will share software and tools via public repositories and scientific workshops, enabling global researchers to utilize and build upon the project's methods.
For families affected by Alzheimer's disease, the ultimate goal is clear: to develop more accurate tools for distinguishing different types of dementia and identifying the most effective therapies for individual patients. By combining large-scale data with advanced AI, AI4AD2 strives to bring personalized medicine closer to reality for one of the world's most challenging neurological diseases. This project represents a significant step forward in our understanding and treatment of Alzheimer's, offering hope for a future where personalized care becomes a reality for those affected by this devastating condition.
