The study, published in the Geriatrics journal, used driving data, demographic information, and machine learning to predict — with 88% accuracy — whether an individual has MCI or dementia. Now, driving data on its own is not the most accurate measure of whether someone is experiencing these kinds of cognitive impairments. Driving data alone can only predict this with 66% accuracy. However, when it’s coupled with demographic data — such as age, race, gender, and education level — the accuracy rises to 88%.
Predictors of MCI or dementia
Age was the leading predictor of MCI or dementia, but a number of driving characteristics also proved to be good indicators. According to the study, this includes things like the number of trips within 15 miles of home, the length of trips, and the number of hard braking events can also be signs. To come to these conclusions, the researchers used data from a long-term study called LongROAD, which has tracked nearly 3,000 drivers for four years. However, the study says that not many individuals with MCI or dementia were found in the LongROAD dataset. So while the researchers’ model has managed to predict cognitive impairment with some reliability, it requires much more data before it can be considered robust. Researchers of this latest study suggest that their machine learning model could be incorporated into a smartphone app or vehicle systems to monitor driving behavior and suggest if the driver is experiencing MCI or dementia. If you want to read the full study, check it out for yourself here. HT – New Atlas Do EVs excite your electrons? Do ebikes get your wheels spinning? Do self-driving cars get you all charged up? Then you need the weekly SHIFT newsletter in your life. Click here to sign up.
