Mon. Dec 23rd, 2024
Biosensor Can Predict Aggression In Autism 3 Minutes In Advance

Matthew Goodwin wears a biosensor on his wrist.
Photo by Matthew Moduno/Northeastern University

It happens “suddenly”.

These are words frequently uttered by caregivers of people with severe autism who become aggressive and violent on a daily basis.

But what if such actions were predictable? How many 911 calls, injuries, or volatile situations could have been avoided?

Professor Matthew Goodwin of Northeastern University recently demonstrated that aggressive behavior in young people with severe autism can be predicted three minutes in advance with 80% accuracy using biosensor data and machine learning. The research was published in JAMA Network Open.

Biosensors can detect specific physiological changes that may indicate potential aggressive behavior.

“Three minutes is plenty of time to do something,” says Goodwin, an interdisciplinary professor in Northeastern University’s Boob College of Health Sciences and Cooley College of Computer Science. “If caregivers knew before[an incident]was going to happen, they would stop whatever they were doing, keep an eye on their child, and rearrange the environment to make it safer. You can try to solve the problem.” It is an escalation strategy, or preemptive intervention, in which aggression is not the only means of meeting the child’s needs. ”

Matthew Goodwin sits at his computer and monitors the results.
Professor Matthew Goodwin, jointly appointed to Bouvet College of Health Sciences and Cooley College of Computer Science; He is working in a laboratory in the basement of Robinson Hall on January 16, 2024.Photo: Matthew Moduno/Northeastern University

According to the Centers for Disease Control and Prevention, 1 Currently, 36 children under the age of 8 have been diagnosed with autism., and there are 78 million people with autism worldwide. Of these 78 million people, 27 percent are thought to have severe autism. This is a condition in which a person has an IQ below 50, has limited or no language, and frequently engages in aggressive behavior that is harmful to themselves, requiring 24-hour care. , another person or property.

But this aggression is very difficult to predict, Goodwin said, and caregivers often report that it happens “suddenly” and that they are unprepared. After all, most people with severe autism are unable to express their emotions clearly, and many are emotionally flat, or “mostly have a poker face,” Goodwin said. says.

“If you don’t know when your child is in distress or suddenly exhibits aggressive behavior, that means crisis management and safety management are your only options when you realize there’s a problem.” says Goodwin. . “This is not a teaching or learning opportunity. It’s a ‘hair on fire’ situation. Stop whatever we’re doing and try to prevent injury or destruction.”

The unpredictability of such explosions puts caregivers, patients, and others at risk.

In fact, Goodwin points out that staff who specialize in caring for children with severe autism experience more injuries, file more insurance claims, and have higher burnout rates.

When such aggressive incidents occur, caregivers often resort to calling 911, visiting emergency rooms, or admitting patients to psychiatric facilities, but Goodwin said care It is said that there are social and economic costs.

For all of these reasons, caregivers and children with severe autism are often isolated at home.

“This is not a path to independence,” Goodwin said.

But Goodwin recognizes that although the signs may not be immediately visible, a child with severe autism’s body still needs to prepare for aggressive behavior.

“When a child begins to hit, punch, kick, or tear up the environment, the body must make preparatory physiological changes to support that increased physical demand. ” Goodwin said.

The question was how to identify these formulations. One answer is biosensors.

While at the MIT Media Lab, Goodwin worked on products that are the predecessors to today’s Fitbits and Apple Watches, and developed products that track “peripheral autonomic nerve signals,” such as heart rate and electrodermal activity (often referred to as “emotional sweating”). We have been waiting for biosensor technology to advance enough to measure “. ,”) the surface temperature of your skin, how much you’re moving, etc. The biosensor also needed to be wireless, off-the-shelf (so not custom-built for research), and wearable in any environment.

That technology is now available.

With funding from the Simmons Foundation, the Nancy Lurie Marks Family Foundation, and the Department of Defense, Goodwin will provide watch-like biosensors to youth with severe autism in four inpatient psychiatric hospitals. did. He also provided staff with a custom mobile app to report the frequency and duration of aggression that involves harm to patients, others, or property.

Goodwin received data from a total of 70 youth, including 497 hours of biosensor and behavioral data, including 6,665 instances of aggressive behavior.

“This is where the real work begins,” Goodwin says.

Goodwin and his team use machine learning and artificial intelligence to analyze the data and discover how much data is needed from the recent past, how far into the future, and how far into the future to predict behavior. We determined whether the prediction could be made with a certain degree of accuracy.

Goodwin said his team’s previous study, conducted at a single inpatient facility with 20 young people with autism, predicted the onset of aggression toward others one minute in advance with 70% accuracy. It points out that it did.

Using more data, the current study was able to predict aggression toward others, self-harm, and meltdowns three minutes in advance with 80% accuracy.

Mr. Goodwin then said, under the supervision of clinical staff trained to determine “the functions of an individual’s aggressive behavior,” such as escaping or avoiding demands or gaining access to preferred objects or activities. , has applied his method in an outpatient setting.

He believes that combining this information with biosensor data will allow clinicians to develop customized intervention plans that can be implemented at home, reducing families’ dependence on emergency room and hospital professionals. I hope that it will be possible.

He is also working on cloud computing, which can analyze biosensor data in real time. This allows parents to be alerted via text message before a violent incident is likely to occur, allowing parents to use the app when they need support without having to constantly monitor their children. You will be able to do it. is required.

“The holy grail is to prevent aggressive behavior from occurring in the first place,” Goodwin said. “Even if you can’t do that, it’s worth being prepared to manage it. Even if you don’t manage it perfectly, you’re not completely caught off guard, so you learn something.” It’s more likely.” What triggered it in the first place, and how can you prepare for it? ”

Goodwin said other scientists (the anonymized dataset used in the published study is available free of charge to eligible scientists) have investigated situations of domestic violence and bipolar He says people have consulted him about applying his methods to deal with other common health situations, such as predicting incidents for people with disabilities. or post-traumatic stress disorder, suicidal thoughts, or self-harm.

But for now, Goodwin is focusing on children with severe autism and their caregivers, an understudied and underserved segment of the autism population. is focused on.

“Families raising children with severe autism who exhibit aggressive behavior have higher rates of divorce, mental health problems and substance abuse problems, and siblings are more likely to be marginalized by their families and school systems. They don’t have the proper care and attention, and they’re having a hard time trying to help that child get by,” Goodwin said. “This is a ripple effect that affects not just that child, but a much larger system. I want to help.”

Editor’s Pick

Recent stories

News, discoveries and analysis from around the world