Health

Near-perfect Digital Markers for Predicting Dementia in Elderly Drivers

Near-perfect Digital Markers for Predicting Dementia in Elderly Drivers

Dementia is a progressive brain disorder that affects a person’s ability to think, remember, and reason. It can also impact their ability to perform everyday activities, including driving. As people age, the risk of developing dementia increases, and it can be challenging to identify early signs of cognitive decline.

Researchers at Columbia University’s Mailman School of Public Health, Fu Foundation School of Engineering and Applied Science, and Vagelos College of Physicians and Surgeons developed a novel, interpretable, and highly accurate algorithm for predicting mild cognitive impairment and dementia in older drivers using ensemble learning techniques and longitudinal data from a large naturalistic driving study.

Variables generated from data captured by recording devices in the real world are referred to as digital markers. These data could be processed in great detail to measure driving behavior, performance, and tempo-spatial pattern. The findings were reported in the journal Artificial Intelligence in Medicine.

Our new ensemble learning model based on digital markers and basic demographic characteristics could predict mild cognitive impairment and dementia in older drivers with high accuracy.

Sharon Di

For selecting predictive variables in the dataset, the researchers used an interaction-based classification method. This learning model outperformed traditional machine learning models such as logistic regression and random forests in predicting mild cognitive impairment and dementia, outperforming a statistical technique widely used in AI for classifying disease status.

“Our new ensemble learning model based on digital markers and basic demographic characteristics could predict mild cognitive impairment and dementia in older drivers with high accuracy,” said Sharon Di, lead author of the study and associate professor of civil engineering and engineering mechanics at Columbia Engineering.

The investigators constructed 200 variable modules using the naturalistic driving data on the driver, the vehicle, and the environment captured by in-vehicle recording devices for 2977 drivers participating in the Longitudinal Research on Aging Drivers (LongROAD) project, a prospective cohort study conducted in five sites across the contiguous United States and sponsored by the AAA Foundation for Traffic Safety.

Digital markers near-perfect for predicting dementia in older drivers
Digital markers near-perfect for predicting dementia in older drivers

At the time of enrollment, the participants were active drivers aged 65-79 years who were cognitively intact. Data used in this study came from the first three years of follow-up, spanning from August 2015 through March 2019. During the follow-up, 36 participants were diagnosed with mild cognitive impairment, 8 with Alzheimer’s disease, and 17 with other or unspecified dementia.

In a series of computer modeling experiments, the researchers discovered that the new ensemble learning model is 6-10% more accurate than random forests and logistic regression models in predicting mild cognitive impairment and dementia. The right-to-left turn ratio and the number of hard braking events (defined as maneuvers with deceleration rates of ≥ 0.4 g) are the two most influential driving variables. “As people get older, they make fewer left turns and more right turns because left turns are riskier,” Di observed.

“In the United States, approximately 85 percent of older adults are licensed drivers. Driving, as the most preferred mode of personal transportation, is critical to maintaining independence, self-control, social connection, and quality of life. Driving a car safely necessitates critical cognitive and physical abilities. Our findings suggest that digital markers embedded in routinely collected driving data can be used as valid and reliable artificial intelligence for predicting mild cognitive impairment and dementia using innovative machine learning techniques” Guohua Li, MD, DrPH, senior author and professor of epidemiology and anesthesiology at Columbia Mailman School of Public Health and Vagelos College of Physicians and Surgeons.

“Early detection of mild cognitive impairment and dementia could lead to timely evaluation, diagnosis, and interventions, which are especially salient in the absence of effective therapeutics.”