Behaviors behind the wheel such as the number of trips that someone takes within a radius of up to 25 kilometers from home, the duration of runs and the amount of sudden brakes performed in the period can help diagnose Mild Cognitive Impairment (CCL) and dementia – the decline in cognitive ability associated with loss of the ability to perform common tasks. This is the result of a study already reviewed, published at the end of April in the scientific journal Geriatrics.
The analysis considered, in addition to data obtained while driving vehicles, demographic information such as age, race and ethnicity, sex and education level of the participants. With the use of machine learning it was possible to predict – with an accuracy of 88% – whether an individual had CCL or dementia.
Advanced age and certain behaviors behind the wheel may indicate the beginning of cognitive impairment.Source: Pixabay
But it is not yet time to frame that friend or uncle as a barber: driving data, by itself, is not the most accurate measure to know if someone is facing some type of cognitive impairment. Driving data alone can predict symptoms with just 66% accuracy. When combined with demographic data, the accuracy increases to 88%.
How it is possible to predict CCL and dementia
Although advanced age is the main factor in predicting CCL or dementia, a number of characteristics in the direction of vehicles have also proved to be good indicators. The new study was based on data obtained from another long-term research – the LongROAD – which filmed about 3,000 drivers driving for four years. With these data, the researchers created 29 variables that consider behavior, use of space and performance in the direction.
The team created a machine learning algorithm to predict dementia.Source: Pixabay
The study was also based on other previous research, in which elderly people were classified according to the levels of amyloid deposition – insoluble protein related to dementia – in the brain. The history of involvement in accidents and violation of traffic rules was highest among participants with higher amyloid deposits – they were also worse in tests on the road and parking.
The new research claims, however, that not many individuals with MCI or dementia were found in the data set obtained by LongROAD. And although the model created for the research was able to predict cognitive impairment with some reliability, more data will need to be incorporated before considering the system robust and ready for commercial use.
Seniors and management
The new study says that while driving allows older adults to have their mobility needs met and remain independent, typical age-related declines – such as medical conditions and medication side effects – can compromise vehicle driving skills, increasing the risk of driving. accidents. He adds that atypical changes in driving behavior may indicate an initial decline in cognitive functions.
Driving can assist in independence during old age.Source: Pixabay
The essay lists the main changes reported in older drivers with CCL or early stage dementia: decline in driving performance, increasing the incidence of facts such as getting lost; driving test failures; reduction in the capacity of spatial navigation and atypical driving behavior, such as the reduction of episodes behind the wheel – avoiding night driving and the time of the rush, for example.
The study team suggests that, in the future, it will be possible to incorporate its machine learning algorithm into a smartphone or in vehicle operating systems to monitor the behavior of individuals in driving. Thus, the application would notify the driver if he or she was showing behaviors characteristic of CCL or dementia while driving.