Using AI to understand mental health

The Canadian Press
Liam Casey

TORONTO–Sean Hill believes the mental health crisis can be improved by harnessing big data.
The Toronto researcher says harvesting all the available information in the field–whether it’s collected by clinicians or trapped inside patients’ genes and neurons–will help answer the big questions about mental disorders. Mental health research, he says, needs to be transformed into a data-driven science.
“Right now mental health disorders are defined in terms of symptoms and we want to be defining them based on biological mechanisms,” Hill said. “That’s the Holy Grail of where we want to go.”
And that’s what the Krembil Centre for Neuroinformatics, officially opened today at the Centre for Addiction and Mental Health (CAMH) in Toronto, promises to do. One of the goals of the new research hub is to combine machine learning and data science, and apply it to neuroscience.
The centre’s team is made up of researchers from Canada, the U.S., the U.K., Israel and Romania.
The scientists will focus on several areas, including computational genomics that will analyze the relationship between genes, cells and circuits in the brain, said Hill, who will lead the team.
Another group will build simulations of brain circuits, while another will focus on relating genes to brain anatomy.
And perhaps the team’s most ambitious project is “whole-person modelling.”
That’s where one team will try to make sense of all the data the centre is collecting, from demographic to brain imaging to genomic information along with an individual’s exercise and sleep patterns.
“How do we take into account all of those individual aspects and build a computational model that can help us understand risk factors and best approaches to treatment?” Hill asked.
He believes the project wouldn’t be possible outside a hospital setting.
“There is just too many challenges around the access and governance of the data and you have to have a very tight collaboration with clinicians anyway,” Hill said. “Being within the same hospital really facilitates that.”
The team has already begun collecting data.
Researchers are working with clinicians who specialize in major depressive disorders to track patients’ progress. Patients now use tablets to fill out forms before each visit that help understand how the individual is doing compared to the previous visit, how they’re responding to medication or therapy and the like.
“We provide a customized digital dashboard to the clinicians so they have a tool of a deeper view into the progress and trajectory of the patients and use that in their treatment,” Hill said.
“What we really want to do is increase the degree of personalization of care.”