Can Artificial Intelligence Boost Mental Healthcare Accessibility? 

According to MIT and Massachusetts General Hospital researchers, artificial intelligence could improve the accessibility of mental healthcare.

The investigation was led by MIT Professor of Media Arts and Sciences Rosalind Picard and Associate Director of the Depression Clinical and Research Program at Massachusetts General Hospital Paola Pedrelli.

“It's been very, very clear that there are a number of barriers for patients with mental health disorders to accessing and receiving adequate care,” Pedrelli, who has been a clinician and researcher in psychology for 15 years, said in a press release

Those barriers can include figuring out when and where to seek help, finding a nearby provider who is taking patients, and obtaining financial resources and transportation to appointments.

To address health barriers Picard and Pedrelli have been working collaboratively for over five years to create machine-learning algorithms that can assist in diagnosing and monitoring symptom changes in individuals with major depressive disorders.

To conduct the study, the research team recruited MGH participants with major depressive disorders who have recently changed their treatment. Thus far, 43 participants have enrolled in the study.

Using smartphones and other wearable devices, the research team can gather detailed data on participants’ temperature, heart rate, activity levels, socialization, personal assessment of depression, sleep patterns, and more.

“We put all of that data we collected from the wearable and smartphone into our machine-learning algorithm, and we try to see how well the machine learning predicts the labels given by the doctors,” Picard said. “Right now, we are quite good at predicting those labels.”

According to researchers, their goal is to develop machine-learning algorithms that can intake large amounts of data and identify individuals that may be struggling with their mental health. The hope is that the algorithms will provide physicians and patients with useful information regarding an individual’s disease trajectory and effective treatment methods.

But developing effective machine-learning algorithms and designing tools that will also empower users poses a unique challenge.

“The question we’re really focusing on now is, once you have the machine-learning algorithms, how is that going to help people?” Picard explained.

The research team is reviewing how the machine-learning algorithms could present their findings to users through new devices, smartphone apps, and so on. According to researchers, allowing participants to view their data could encourage them to engage in certain behaviors that improve their well-being.

However, if implemented incorrectly, the technology could have a negative impact. If the app indicated that someone is heading toward a deep depression, that could be discouraging information that leads to further negative emotions. Pedrelli and Picard are including real users in the design process to develop a tool that is helpful, not harmful.

“What could be effective is a tool that could tell an individual ‘The reason you’re feeling down might be the data related to your sleep has changed, and the data relate[d]to your social activity, and you haven't had any time with your friends, your physical activity has been cut down. The recommendation is that you find a way to increase those things,’” Picard said.

Additionally, the team is prioritizing data privacy and informed consent.

AI and machine-learning algorithms have the unique ability to make connections and identify patterns in large datasets.

“I think there's a real compelling case to be made for technology helping people be smarter about people,” Picard said.