AI-powered rectangular box can detect Alzheimer's beforehand
Alzheimer’s, a progressive disease that demolishes brain’s memory and its regular functioning as well. There isn't any single test till date to diagnose this disease and brain scans alone can’t determine whether the person is possessed by it.
Currently, the physician believes that a person is affected by Alzheimer’s by based the reports of the family members about the behavioural tendencies and observations of the past medical history. AI combined with Machine Learning algorithms might now be able to change this situation.
David Graham, a patient who is suspected to have an Alzheimer’s disease hasn’t been diagnosed yet, the researchers at MIT have been monitoring him along with four other patients, who are been already diagnosed with the disease. Graham has an AI-powered rectangular box attached to his room’s wall, which is an assisted living facility in Marlborough, Massachusetts.
The AI-powered box on his wall tracks the moments of Graham, it knows when he gets out of his bed, gets dressed, goes to the bathroom or whenever he walks to his window. It can also track his sleeping habits of whether he’s sleeping or has already fallen asleep. It does this by generating low-power signals to monitor/track his gait speed, sleep patterns, location, and even breathing patterns as well. All this tracked information will be stored in the cloud, where machine learning algorithms come into action to study is thousands of movements and patterns every day.
According to MIT, these rectangular boxes are part of an experimental process, through which researchers can find, track and understand the symptoms of Alzheimer’s. Researchers say that tracking symptoms in the early stages of this disease are highly unlikely, but, with the help of AI, it might be possible. This AI-based recognition can help identify the patients, who are at a risk of developing the most severe form of the Alzheimer’s disease.
Devices with such capabilities can be placed in any home and facilities where continuous monitoring of the patients is necessary for those at risk. These devices can also be used by drug manufacturers to collect data of the experimental drugs, which are being helpful on the patients in the course of the treatment.
According to the report, Dina Kabati, the director of the MIT Wireless Center and her team at MIT’s Computer Science and Artificial Intelligence laboratory had developed this instrument to monitor people without the need of a wearable tracking device. “This is completely passive. A patient doesn’t need to put sensors on their body or do anything specific, and it’s far less intrusive than a video camera,” she says.
Vahia, a geriatric psychiatrist at McLean Hospital and Harvard Medical School, along with the technology’s inventors at MIT are running a small pilot study of the device.
The device works based on a wireless signal, which when being sent reflects of everything in a 30-foot radius that includes human bodies as well. As they had this device equipped in Graham’s room, they could track each and every moment of Graham including his breathing pattern (a signal reflects according to this pattern).
Katabi developed a machine learning algorithm to track and analyse all these minute reflections that include all the complex movements of Graham on a day-to-day basis. “As you teach it more and more, the machine learns, and the next time it sees a pattern, even if it’s too complex for a human to abstract that pattern, the machine recognizes that pattern,” Katabi says.
The AI on this device can pick out the patterns which signify things like depression, agitation and sleep disturbances. It can also pick up behavioural changes during the day which the patients tend to repeat. These all come under standard symptoms of Alzheimer’s disease. “If you can catch these deviations early, you will be able to anticipate them and help manage them,” Vahia says.
Marilyn Miller, who directs AI research in Alzheimer’s thinks that AI could be used to diagnose and predict Alzheimer’s in patients in as soon as five years from now. But she says large amounts of this reliable and accurate data needs to be collected to analyse these Machine Learning based algorithms.