<< Back

2019 Alzheimer's Association Clinician Scientist Fellowship (AACSF)

Optimizing prediction of clinical progression in MCI population

Can cutting-edge computer science and statistical techniques correctly predict which individuals are at higher risk for cognitive decline and dementia?

Ali Ezzati, M.D.
Albert Einstein College of Medicine, Inc.
Bronx, NY - United States


The availability of vast data sets of information for each individual in large studies as well as recent technological advances in the field of computer science called artificial intelligence, have enabled scientists to apply sophisticated computational tools to biomedical research.  Researchers are trying to identify individuals who are at higher risk for Alzheimer’s dementia before they start showing the clinical symptoms – changes in memory, thinking and reasoning. Dr. Ali Ezzati proposes to use cutting-edge statistical and computer science techniques to better determine risk for cognitive decline and dementia in individuals based on their individual information.

Research Plan

To conduct their study, Dr. Ezzati and colleagues will use data from 1) the Alzheimer’s Disease Neuroimaging Initiative (ADNI) - a collaborative study by researchers all across the U.S. to help understand the progression of Mild Cognitive Impairment (MCI) to Alzheimer’s dementia and from the 2) Einstein Aging Study (EAS). Both studies are among the largest and longest population studies related to aging and dementia. ADNI researchers collect biological marker (biomarker) data that includes brain scans, genetics, blood and cerebrospinal fluid (a biological fluid that surrounds the brain and spinal cord) from a large number of people over the last fifteen years.
Dr. Ezzati and his team will apply advanced statistical methods on both the datasets (comprising of individual measurements of brain scans, cognitive tests, biological markers etc.) to identify the optimal and most helpful indicators of high dementia risk. Subsequently, the researchers will apply this information to their advanced computer science technique (called “machine learning”) to accurately predict which subgroup of older adults with MCI maybe at higher risk for cognitive decline and may progress to Alzheimer’s during the first 5 years of follow-up.


If successful, the techniques used in this study could be used for an improved and accurate early detection of individuals at high risk of cognitive decline and dementia. Families facing Alzheimer’s now and in the future will benefit greatly from early detection, allowing for important care and planning. Furthermore, when we have new therapies, we will be in a better position to know who needs treatment at the earliest time point. 

Back to Top