Predicting dementia risk using neuroimaging and cognitive assessment.

Area of Science:

• Neuroscience
• Gerontology
• Medical Imaging

Background:

• Dementia affects over 55 million globally, with numbers projected to double.
• Current dementia risk assessments often fall short, necessitating improved early detection methods.
• Integrating structural brain imaging with cognitive assessments may enhance dementia risk prediction accuracy.

Purpose of the Study:

• To evaluate the efficacy of combining structural brain imaging and brief cognitive assessments for predicting dementia risk.
• To compare the predictive power of single-visit versus longitudinal modeling approaches.
• To identify key neuroimaging and cognitive markers associated with dementia risk.

Main Methods:

• Utilized data from 312 older adults from the KU Alzheimer's Disease Center cohort.
• Employed cognitive testing and MRI scans measuring hippocampal volume, gray matter, and Alzheimer's disease signature regions.
• Assessed depressive symptoms using the Geriatric Depression Scale (GDS) and analyzed both single-visit and longitudinal data.

Main Results:

• Multimodal models incorporating neuroimaging significantly outperformed those relying solely on demographics or cognitive scores.
• The optimal model, combining imaging and cognitive data, achieved 77.6% accuracy in predicting dementia status.
• Reduced hippocampal volume, lower gray matter, and higher GDS scores were identified as key predictors, suggesting an interaction between depression and neurodegeneration.

Conclusions:

• A compact, multimodal approach using standard MRI and brief cognitive tests can generate clinically relevant individualized dementia risk profiles.
• This method provides a scalable pathway for early intervention, clinical trial enrollment, and personalized care planning.
• Future research will focus on diverse population validation and integration of fluid biomarkers to refine dementia risk prediction.