The Steinberg Centre for Simulation and Interactive Learning is supporting Aifred Health as they conduct ease of use pre-clinical trials for their predictive AI depression treatment app
Mr. Roberts is an 82-year-old retired man who is visiting his family doctor because he isn’t quite feeling like himself lately. He’s having trouble falling asleep, isn’t enjoying his usual activities, and is feeling anxious. Before his appointment, he filled in a short online questionnaire which the doctor will use to help ensure that he is receiving the best treatment for his condition.
This is one of three simulated scenarios that recently took place at McGill’s Steinberg Centre for Simulation and Interactive Learning (SCSIL) with standardized patients—actors who have been trained to accurately and consistently recreate the parameters of an actual patient at a specific point in time.
The scenarios are based on real patient interactions and were developed in partnership with Aifred Health, a McGill-based start-up that has created a clinical decision aid for family doctors and psychiatrists who are treating depression.
Depression is a common disorder which can be difficult to treat. Aifred Health has devised a solution that brings personalized medicine to psychiatry, making use of innovative and powerful machine learning techniques to help physicians choose the most effective treatments for a given patient—whether that be a specific medication, psychotherapy, or other possible treatment—based on the individual patient’s profile. This is the first time that this type of device has been implemented in mental health in Canada.
Working in collaboration with the SCSIL team, the multidisciplinary Aifred Health team conducted a series of ease of use pre-clinical trials to assess their technology. Over the last month, psychiatry and family medicine residents and staff physicians have participated in this study by using the app under simulated conditions, with the goal of garnering valuable feedback on the tool and evaluating how it impacts the patient-clinician relationship. They wanted to see if the clinicians find the tool easy to use, if they generally agree or disagree with the model’s treatment recommendations, and if they find the extra features helpful in their clinical practice.
“We strongly believe in the potential for artificial intelligence to enhance, but never replace physician decision-making. Following this principle, the model must be user-friendly and provide clinicians with features they want and need,” explains Dr. David Benrimoh, Chief Science Officer of Aifred Health. “Using AI, we should be able to learn from thousands of patients to predict which treatment will be optimal for which patient.”
October 9 2019