Alzheimer's Early Detection: Integrating Omics and Imaging with AI
Overview
At the Alzheimer’s Association International Conference 2019, a collaborative research team, including experts from Ariana Pharma and several prominent academic institutions, presented groundbreaking findings on combining omics and imaging data for early detection of Alzheimer’s disease. This study highlighted the use of explainable Artificial Intelligence to identify genomic biomarkers that are pivotal in the early, presymptomatic identification of Alzheimer’s, such as symptoms in individuals showing subjective memory complaints (SMC) yet displaying unimpaired cognition and memory.
Impact
• The integration of genomic data with brain imaging techniques such as functional Magnetic Resonance Imaging (fMRI) through AI technologies allows for the identification of early biomarkers for Alzheimer’s disease, significantly advancing the potential for early diagnosis and intervention.
• The research supports the critical need for comprehensive data analysis in understanding the complex interplay between genetic factors and brain structure evolution across aging, potentially revolutionizing Alzheimer’s disease management.
Objectives
• To identify genomic markers that correlate with changes in neuroimaging, providing early indicators of Alzheimer’s disease in preclinical subjects.
• To utilize combined omics and imaging data to refine predictive models for neurodegeneration, enhancing the accuracy of early-stage Alzheimer’s disease detection.
Method
The methodology involved systematic genomic profiling of subjects combined with detailed neuroimaging, including PET scans and fMRI. By applying KEM® Artificial Intelligence technology, the team generated, explored, and ranked numerous genomic-imaging associations, leading to the identification of critical biomarkers linked to early signs of Alzheimer’s disease.
Results
• The research identified specific genomic variants linked to neuroimaging features associated with Alzheimer’s disease in other clinical context. For example, a significant association was found between a homozygous variant in the COG6 gene and decreased activity in the Orbito Frontal Cortex, indicating early neurodegenerative changes.
• These findings demonstrate the efficacy of integrating multiple data types to uncover new insights into the pathophysiology of Alzheimer’s disease, setting a foundation for further research and potential clinical applications.
