Early Signs of Dementia Can Be Detected by Tracking Driving Behaviors

A new study published in Geriatrics used machine learning (ML) algorithms that can analyze naturalistic driving data and identify mild cognitive impairment (MCI) and dementia in drivers, reports New Atlas . The researchers say such methods could be used to detect early signs of dementia using either a smartphone application or devices incorporated into car software systems. The researchers employed data from the Longitudinal Research on Aging Drivers study, which monitored nearly 3,000 older drivers for up to four years, diagnosing 33 subjects with MCI and 31 with dementia. A series of ML models were trained on this data. "Based on variables derived from the naturalistic driving data and basic demographic characteristics, such as age, sex, race/ethnicity, and education level, we could predict mild cognitive impairment and dementia with 88 percent accuracy," said Columbia University's Sharon Di. While age was the lead variable for detecting MCI or dementia, a number of driving factors closely followed, including "the percentage of trips traveled within 15 miles 24 km of home . . . the length of trips starting and ending at home, minutes per trip, and number of hard braking events with deceleration rates equal or greater than 0.35 g." The models could still predict drivers with MCI or dementia with 66 percent accuracy from driving variables alone.