Title : MRI biomarkers and predictive modeling of disease severity

This oral presentation highlights two complementary research projects that integrate neuroimaging biomarkers, demographic determinants, and gene–environment interactions to advance methodological approaches in Alzheimer’s disease research and predictive analytics.

Part I: MRI Biomarkers and Environmental Risk in ADNI

The first component examines structural MRI biomarkers within the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset. Using student-led segmentation workflows and volumetric extraction of hippocampal and cortical regions, the analysis investigates relationships among neuroanatomical atrophy patterns, environmental exposures, and metabolically relevant genes such as EPHX1 and PAH. Preliminary models demonstrate significant associations between hippocampal thickness and exposure-linked biological pathways, suggesting a measurable contribution of environmental risk interacting with genetic susceptibility. These findings reinforce the importance of multidimensional imaging-genetics approaches in understanding early neurodegenerative change.

Part II: Predictive Modeling of Disease Severity Using Demographic and Gene–Environment Indicators

The second component centers on building predictive algorithms to estimate disease severity by integrating demographic variables (age, sex, race/ethnicity, comorbidity profiles) with genetic markers and potential environmental interactions. Using generalized estimating equations (GEE), generalized linear mixed-effects models (GLMM), and machine learning pipelines, the analytic framework identifies key predictors that significantly influence severity trajectories. Early results indicate strong predictive contributions from demographic profiles and specific gene–environment interactions, enabling development of a scalable analytic model for disease-risk stratification. This provides a foundation for more personalized, population-level decision support tools in public health and clinical research.

Together, these two projects demonstrate an innovative training pipeline and a cross- disciplinary research model that merges neuroimaging analytics with advanced predictive modeling to explore the biological, demographic, and environmental architecture of disease severity in Alzheimer’s and related disorders.

Dr. Tia Warrick, DHSc, MPH, CCRA, is an epidemiologist and Director of Public Health at Juniata College, a senior clinical research project manager, and principal investigator of integrative research initiatives that merge neuroimaging, machine learning, and public health analytics. Her work bridges academic and industry methodologies to advance imaging–genetics research and predictive modeling for disease severity.