Skip to main content

Using AI to ease troubled minds: a remarkable search to improve treatment for depression


Nearly  5 per cent of Canadians suffer from depression at some point in their lives; many find the road to feeling better is a rough one. While there are more than a dozen drugs known to be effective in treating depression, roughly two-thirds of patients have to try up to four treatment courses before finding one that helps. It’s a process that can take months, and in this period to “wait and see what works”, many patients continue to struggle.

Dr. David Benrimoh, Chief Science Officer for Montreal start-up Aifred Health, is continuing to hone its’ artificial intelligence (AI) tool to try and break this cycle. Aifred uses deep learning to find the links between patient characteristics and symptoms and which drugs have worked for them.

“Basically, the system (Aifred) learns about patterns that tell us how [other] patients have responded to treatments in the past, which then allows us to calculate for each patient what the likely response to different treatments would be, so we can actually run a single patient across all the treatments that we have in the system and report the probabilities of success for each one.” 

That percentage rating of likely effectiveness, plus information on guidelines and other relevant details are returned to the clinician, who makes the final decision on what treatment to prescribe.

Learn more about Dr. Benrimoh and how Aifred Health is disrupting the “guess and check” approach to treatment in our 2019 podcast

Dr. Benrimoh says the $25,000 Joule Innovation grant his team received in 2018 was critical to growing Aifred. It allowed the start-up to retain the staff it needed to continue building the product and to grow enough as a company to secure more stable financing.

The grant also brought important opportunities for mentoring and networking.  Aifred team members were able to meet people at different hospitals who would be interested in the decision support tool.  Today, the system is being used for studies or pilots at four clinical sites in Canada.

The company recently received another welcome reward for its progress, making it as a semi-finalist for the IBM Watson AI XPRIZE, a 4-year competition that launched the company. Before that, in November 2019, they won a “milestone” prize of $35,000.  While the team is still waiting to hear if they have won the competition, which will be announced in April 2020, their focus remains on getting the decision support tool into the hands of more physicians.

Dr. Benrimoh points out that today, roughly 90 per cent of treatment for depression is done by family practice physicians, who don’t have as much training and experience as specialists with different treatments.

“A lot of doctors we’ve talked to said they were very happy to have something that reminded them of all the options there are and how to use them, and when they might be appropriate for different patients,” explains Dr. Benrimoh.

The next challenge ― and the toughest ― is to move Aifred beyond the testing phase. Widespread use is the goal, and Dr. Benrimoh says this vision for Aifred is what has kept the team moving forward.

Tweetable: “Rolling out our solution to millions of patients, to reduce costs and suffering — and then being able to move from depression to other therapy areas like schizophrenia, anxiety and bipolar disorder, this is our goal and as long as it is possible, we will keep working to get there.”


This material is for informational purposes only. It is not intended to be a substitute for professional medical advice and should not be relied on as health or personal advice. The opinions stated by the authors are made in a personal capacity and do not necessarily reflect those of the Canadian Medical Association and its subsidiaries including Joule.  Feel passionate about physician-led innovation? Please connect with us at