According to the World Health Organization, the number of people affected by Alzheimer’s and dementia will triple over the next 30 years, and up to 60 percent of them will wander and become lost at least once.
A new study by George Mason University’s College of Health and Human Services analyzed data from GPS trackers to predict if individuals with dementia are wandering. They used machine learning methods to identify routine movement patterns and other patterns that may indicate wandering.
“When individuals with Alzheimer’s or dementia wander, it can be dangerous and cause their loved ones great concern,” said Janusz Wojtusiak, associate professor and director of health informatics at the college.
“Until now, there hasn’t been the ability to predict the movements of these individuals as they are wandering,” Wojtusiak said. “We’re proud to take the lead on potentially finding a long-sought solution to this puzzle.”
Individuals’ data were collected from the GPS SmartSole, a tracking device developed by GTX Corp that includes GPS and GSM (Global System for Mobile communications) units embedded in shoe soles to provide real-time geolocation data for wearers. A sample of 338 GPS trackers with at least 14 days of data was used to study wandering habits in elderly people.
Wojtusiak’s research predicted locations and routes of movement with high accuracy, which allows the opportunity to detect unusual movements that may correspond to wandering.
“This is truly groundbreaking data that offers the potential to save lives and dramatically reduce the costs of high-resource rescue operations,” said Andrew Carle, adjunct professor and founding director of the Senior Housing Administration Program at Mason.