For information and resources regarding returning to campus and COVID-19 please visit the university COVID-19 website
Health Sciences professor Matthew Goodwin is on a roll. In the past month, he has published two papers, one on the use of wearable devices to predict aggressive outbursts among youth with autism, and another on the use of a robotic teddy bear to ease anxiety in hospitalized children.
Both harness the power of technology to improve health outcomes among vulnerable populations. This is right up Goodwin’s alley as a founder of Northeastern’s doctoral program in Personal Health Informatics—the only PhD program of its kind in the country. Because of the cross-disciplinary nature of his field, Goodwin has joint appointments in the Bouvé College of Health Sciences and the Khoury College of Computer Sciences.
“Personal health informatics is all about empowerment,” said Goodwin. “It’s about helping patients take control of their own health and stay out of the hospital, which makes healthcare less expensive, more sustainable, and produces better outcomes.”
Several years ago, Goodwin was on the team that developed the wearable sensor used in both studies. The sensor, which looks like an ordinary digital wristwatch, records peripheral physiological responses of study participants. In the past, that data would have to be collected via electrodes attached to the person’s fingers and chest with wires connected to a computer. This made it impossible to study a patient’s response while freely moving in their natural environment. Goodwin’s biosensor, and devices like it, have made the field of personal health informatics possible.
“In the field of autism, we were working with participants who could not tolerate the lab, so we brought the lab to them,” he said.
What if we could predict violent behavior before it happens?
In the case of autism, Goodwin has devised a way to use his biosensor to predict a violent outburst at least a minute before it occurs. This may not sound like much, but a 60-second warning is enough to allow caregivers to provide “just-in-time” interventions designed to prevent physiological distress from escalating into unpredictable aggression.
This could provide an enormous breakthrough in the field, since unpredictable aggression is the biggest single barrier to obtaining community, therapeutic, medical, and educational services for people with autism.
The study, which was published in June in Autism Research, involved 20 youth with autism, 85 percent of whom were only minimally verbal. Goodwin and his colleagues used the sensors to collect physiological and movement data during 69 observational sessions that totaled 87 hours.
His algorithm was able to predict aggressive outbursts with 85 percent accuracy.
“This could ultimately reduce the occurrence, duration, and impact of aggression in youth with ASD, enabling them to more fully participate in their homes, schools, and communities,” he said.
Two Northeastern electrical engineering professors—Stratis Ioannidis and Deniz Erdogmus—contributed to the study, along with researchers from the University of Pittsburg School of Medicine and the Maine Medical Center Research Institute.
It’s blue, it’s fuzzy, and it talks to children when they are most frightened and alone.
Meet Huggable, a robotic teddy bear that talks and plays with hospitalized children who are either waiting for major surgery or battling a serious illness. The children bond so strongly with the bear that they ask it personal questions, tell it jokes, and invite it to come back to meet their families.
For now, the furry robot is controlled remotely by a human who determines its facial expressions, eye movements, and speech. But the authors estimate that they will be able to produce a completely autonomous Huggable in less than three years.
The study, published in the July 2019 issue of Pediatrics, involved 54 patients at Boston Children’s Hospital. They were divided into three groups that interacted with either Huggable, the bear’s computerized avatar on a tablet computer, or a traditional teddy bear.
The authors showed the ultra-cute robotic companion reduced anxiety among hospitalized children, while increasing their communication, joyfulness, and physiological activity. Children who played with Huggable scored better in all these categories compared to those who played with either its computerized avatar or a traditional teddy bear.
“A hospital is a scary place for children who are away from their families for anywhere from one to four weeks,” said Goodwin. “Their anxiety can have a negative impact on their health outcomes.”
In addition to Goodwin, the study was conducted by researchers from the MIT Media Lab and the Harvard faculty at Boston Children’s Hospital. Goodwin and his postdoctoral research associate, James Heathers, oversaw the experimental design and physiological data analysis, while MIT developed the robot and Children’s Hospital provided the clinical expertise and study participants.
While the pilot study was limited by the small sample size, the authors concluded that Huggable introduces “a new way to address the emotional needs of hospitalized children and is deserving of further study.”