AUSTIN — When wearable tech hit the market, a whole new world was opened up for the health and fitness industry. Athletes, teenagers, moms and dads could wear a FitBit or the Apple Watch and unlock an unprecedented amount of data about their own bodies.
The fitness trackers collect health information that reveals trends on the user’s exercise habits, cardiovascular health, menstrual cycles, sleep patterns and more. That data can be used to empower people to make informed decisions about their health.
When the researchers at the University of Illinois Chicago’s Center on Depression and Resilience saw the potential that these wearable technologies could have on helping people seize control over their own health, they began to wonder how this tech could be implemented across other areas of the health spectrum.
“We asked ourselves the question, if we can do this for the human body, why can’t we do this for the human brain?” said Dr. Alex Leow, a UIC associate professor in psychiatry and bioengineering and a researcher at the Center on Depression and Resilience.
That question led the center to launch the Digital Mental Health Initiative, an effort to integrate health tracker tech and artificial intelligence to develop free tools for bringing mental health treatment to vulnerable populations.
The project identified a major void for mental health services on the West Side in particular, exacerbated by the closure of more than half of the city’s mental health clinics in 2012. As the initiative moves from the research phase to implementation, the researchers will place an emphasis on rolling out the technologies in Austin, Lawndale, Garfield Park, Pilsen and Little Village, areas that are medically underserved where patients face disproportionate barriers to high-quality health care.
Patients living with mental and behavioral illnesses in these areas face a unique set of social and economic challenges to accessing care. According to a 2017 survey from Sinai Community Health, 25 percent of adults in Lawndale suffer from Post Traumatic Stress Disorder symptoms, while one in six experience symptoms of depression. But those symptoms often go undiagnosed due to the high cost of care, lack of health insurance, and difficulty finding a provider.
According to Dr. Olusola Ajilore, one of the researchers with the Center on UIC’s Depression and Resilience and an associate professor in psychiatry, a major barrier for people in need of care is a culture of silence around the topic of mental health that stops people from reaching out for help. That stigma is even more prevalent in communities that have been historically cut off from mental health resources, leaving the people with the greatest need alienated from care, Ajilore said.
With the artificial intelligence technologies being developed by Ajilore and Leow, people would be able to access free resources to support their mental health on their own terms, avoiding the stigma that some people feel when reaching out to friends, family, and medical providers.
“We can use technology that’s accessible to most people to get the help which currently isn’t accessible to most people,” Ajilore said.
Ajilore is the lead investigator for the development of DiaBetty, an Amazon Alexa based digital diabetes coach. The application leverages the virtual assistant to help patients manage their diabetes treatments, while also incorporating artificial intelligence to be responsive to a patient’s mental health.
The tech allows Alexa to learn to read speech patterns and the acoustic features of a user’s voice to detect things like stress, mood and emotional state. Stress is an important factor for treating a patient’s diabetes, Dr. Ajilore said, and collecting that type of data can also help patients manage their mental and emotional wellbeing, which can be especially tricky for people living with diabetes.
“Patients who have Type 2 diabetes are twice as likely to develop depression compared to those without diabetes. And if you have depression alongside diabetes, the complications are much worse. … So the presence of depression in the context of diabetes is really important to manage,” Ajilore said.
While DiaBetty alone cannot treat mental illness, Ajilore sees it as an important tool for coordinating care for treating disease. By assessing the emotions of a patient, the artificial intelligence can give advice on how best to self-manage diabetes in the context of a person’s mood and lifestyle.
Meanwhile, Leow is spearheading the development of BiAffect, which is already available at the iPhone App Store. The app is like a mental health fitness tracker that provides data about their cognition based on the user’s keyboard patterns.
No need to worry surveillance: the app doesn’t track what a person types — it tracks how a person types.
Leow found that typing patterns on a person’s smartphone can reveal a lot about their neurological health. By learning about keystroke dynamics like typing speed, frequency of errors, and patterns on social media, the app assesses the user’s mental status and can point out abnormalities to them.
“Nowadays, people are interacting with their personal smartphone as much as they are interacting with another person,” Leow said. Since the phone is often the first and last thing many people interact with on a daily basis, she said that makes it the perfect tool for recording cognitive data.
As the BiAffect app runs in the background of a person’s phone, their keystrokes provide data that Leow’s lab finds to be an effective window into the machinations of a person’s brain. On the app’s dashboard, users can peer into some of the patterns in their typing that can help them to identify thought processes and behavioral symptoms that might indicate underlying disease.
“When the depression symptom rating is higher, then we start seeing more typos being registered by the autocorrect,” Leow said. These types of patterns make the app useful for identifying manic and depressive episodes like those experienced by people living with bipolar disorder. The app is effective at identifying these episodes when typos may become more frequent and erratic, type speed might slow, and a user may be typing at odd hours of the night due to a disrupted sleep schedule.
“So it turns out that, leveraging all this information, we can really construct a model that doesn’t necessarily require a provider to ask a person a lot of questions,” Leow said of BiAffect’s ability to help patients glean information about their own bodies. While the app is already available for free, the lab is working on moving towards implementing the tech in areas where people lack mental health resources like the West Side.
“I feel like now that we have the foundation, the next phase can be about building out and improving our technology across different platforms, and also making sure that it’s ready for a semi-commercial use,” Leow said.
While these apps are not a replacement for a physician, they hold the promise of empowering patients with the information they need to be active participants in maintaining their mental and emotional health.
“The doctors, they really have the knowledge. But the patients are the ones who live every single day with the symptoms,” Leow said. “So if you give them the right information they can leverage that information to learn more about the inner workings of their own brains.”
Pascal Sabino is a Report for America corps member covering Austin, North Lawndale and Garfield Park for Block Club Chicago.
Do stories like this matter to you? Subscribe to Block Club Chicago. Every dime we make funds reporting from Chicago’s neighborhoods.
Comments are closed.