Proving the Business Case for the Internet of Things

AI predicts wandering in Alzheimer’s patients

Steve Rogerson
August 21, 2018

Virginia’s George Mason University’s College of Health & Human Services is using machine learning to predict when people with Alzheimer’s disease may start wandering. It is using asset-tracking technology from California firm GTX.
GTX specialises in wearable GPS human and asset-tracking systems and wandering assistive technology.
The college has received a grant from the Alzheimer's & Related Diseases Research Award Fund to continue its machine learning wandering prediction research. Mason selected the GTX GPS SmartSole and tracking technology platform to conduct the study and continued research, which has the potential to detect quickly and stop episodes of wandering before they can result in a tragedy. This is the first study to link movement patterns to the progression of Alzheimer’s disease.
In June, Mason published a study that explored the possibilities of using machine learning methods and artificial intelligence applied to data from GPS trackers to create models that can predict patterns of movement. Based on the success of the published study, Janusz Wojtusiak, associate professor and director of health informatics at the college, applied for a grant to further the study, which would be for a much larger group of enrolled participants with Alzheimer’s stages three to six.
The study is funded for one year, but will continue with participants as long as possible to collect longitudinal data and, if needed, will replace the SmartSoles over a period of two to three years.
The grant was approved in July and the study, which began immediately, will focus on collecting initial tracking data for people with Alzheimer’s using GPS trackers, along with their medical history and sociodemographic data. The data will then be used to test the feasibility of finding patterns of movement and getting lost.
The participants will be located throughout Virginia and, as part of the study, a social worker will visit participants to show them how to use the GPS devices, do the initial survey and help register online. As part of the grant, Mason will purchase the GPS SmartSoles from GTX and GTX will provide the monitoring services.
“This is so remarkable and truly innovative thinking,” said Patrick Bertagna, GTX CEO. “We began supporting Mason in 2016, with their first small and no budget study, and to see this work evolve to this degree is really impressive. Knowing that our technology continues to prove out its efficacy in what could potentially help millions of seniors afflicted with dementia is such a validation to our core mission statement. We are humbled and very proud to play a small roll in Professor Wojtusiak's incredible forward-thinking and game-changing research.”
In addition to the participants wearing the GPS SmartSoles, Wojtusiak plans to integrate his system with the GTX platform using the displayed maps, highlighting areas in which the models predict a person to be.
“The future of patient monitoring is in advanced technologies such as GTX’s SmartSole, aided by artificial intelligence and prediction,” said Wojtusiak. “By combining these two technologies, we can achieve what has not been possible before. The new study is getting us one step closer to safe care for people with Alzheimer’s and understanding how wandering patterns relate to progression of the disease.”
According to the World Health Organization, the number of people affected will triple from 50 million to 152 million, with an estimated annual cost of $818bn.
“This pilot could produce the ground-breaking data that may have the potential to save lives and dramatically reduce the rising costs of this disease,” said Andrew Carle, adjunct professor and founding director of the programme in senior housing administration at Mason.
The research will be made available to GTX, which will consider deploying the prediction algorithms into its backend monitoring platform, adding another layer of technology and overall value to the company’s proprietary GPS tracking platform and monitoring services.
“Professor Wojtusiak’s team is doing a fantastic job combining advanced machine learning methods with our wander assistive GPS wearable technology,” said Andrew Duncan, GTX director of business development. “This research, combined with AI, could play a major role in finding solutions to complex problems derived from Alzheimer’s, dementia and autism.”
The patented GPS SmartSole is a non-visible GPS tracking device designed to monitor the location of people afflicted with cognitive memory disorders such as Alzheimer’s, dementia, autism and traumatic brain injury, who have a tendency to get lost or wander.
George Mason University is Virginia's largest and most diverse public research university. Located near Washington, DC, Mason enrols 36,000 students from 130 countries and all 50 states.