Gayatri Rajan: Predicting Alzheimer's

“Man’s mind, once stretched by a new idea, never regains its original dimensions.” Oliver Wendell Holmes, Jr.

More than 35 million people worldwide - 5.5 million in the United States alone - have Alzheimer's disease(AD), a condition which causes progressive deterioration of cognitive function. Current predictive models for earlier screenings for AD rely on full input of covariates requiring clinical assessment, such as brain scans, or lifestyle details.

My goal is to build a free application which leverages routine data such as medications prescribed, recent health conditions, and doctor’s notes to detect a higher risk of AD. If a patient’s file contains any recent imaging (CT scans, MRI brain scans, PET scans), these will be also analyzed for feature extraction. The flexibility in covariates will allow for widespread screenings of cognitive decline, leading to opportunities for more aggressive and successful treatment.

Mild cognitive impairment (MCI) is a stage in the transition between age-related cognitive decline and AD. By extracting predictors from training data and isolating the most germane with a recursive feature elimination function, the tool will achieve greater precision. The extracted features will be analyzed via several test algorithms(stratified by patient age). Then, a Clinical Dementia Rating (a clinical measure of the severity of cognitive impairment) will be calculated. Because the tool is free and accessible, this solution can reach doctor’s offices around the world, leading to earlier interventions and more successful treatment plans.

Back to people