Blog
Using AI to Improve Mental Health Research Responsibly: Our New Partnership With Dell, Nvidia, and Brain Canada
We’re working on a solution to one of our biggest challenges: how to use sensitive patient data to advance children’s mental health research while keeping that data completely private
When a child starts showing symptoms of mental health challenges, it takes on average 8 years before they get treatment. That’s a long time for a child to struggle.
Through our Healthy Brain Network, we’ve collected data from over 8,000 children — brain scans, behavioral assessments, clinical evaluations, and more — that researchers around the world have used to better understand how young brains develop.
But here in lies the challenge. We also have hours of video and audio files, detailed notes, and clinical reports that would be incredibly valuable for research. But we can’t share it because it contains private patient information. These are kids who are already vulnerable, and their safety comes first.
Our Approach: Synthetic Data
We’re exploring potentially using artificial intelligence (AI) to create synthetic datasets that appear like real patient data. Think of it as creating realistic practice data that researchers can use to develop better tools, without involving any actual patient information. Researchers could analyze patterns and develop diagnostic tools using this artificial data, then validate their findings on real (but protected) datasets without compromising patient privacy.
To make this work, we need significant computational resources and expertise we don’t have in-house. That’s where Dell Technologies and Nvidia come in. They’re providing the technology and processing power needed to generate high-quality synthetic medical data from complex information like brain scans and behavioral patterns.
Additionally, Brain Canada is contributing significant support to us and our partners at the Center for Addiction and Mental Health, University of Toronto, McGill University, and other world-class institutions to ensure we have the right scientific expertise. This collaboration will help us maintain the rigorous standards needed for eventual use across healthcare systems.
What Success Could Mean
Imagine if researchers around the world were able to safely access and analyze much larger datasets without compromising patient privacy. For families, this could mean faster development of better diagnostic tools and treatments. We could potentially cut down that 8-year wait time, getting kids the help they need sooner when early intervention matters most.
The ripple effects could look like earlier detection of mental health challenges, more personalized treatments, and insights that work across diverse populations beyond the limited groups we typically study.
The Bigger Picture
This work is already opening doors. Hospitals and research centers worldwide are facing the same challenge, and the approaches we’re developing will be available for them to use. We’re creating a blueprint for responsible data sharing that could unlock millions of currently inaccessible medical datasets globally, enabling research at a scale we’ve never achieved before.
The collaboration is proof that complex healthcare challenges can be solved when organizations combine their strengths. We’re seeing what’s possible when clinical expertise, cutting-edge technology, and research infrastructure align around a shared mission.
Moving Forward
Over the next 18 months, we’ll be testing different approaches to synthetic data generation across the various types of medical information we collect. We’re building validation systems to ensure the science is sound. And if this first phase succeeds, we’ll launch data science competitions that engage researchers worldwide.
This partnership is already generating momentum beyond what any of us could achieve alone. We’re not just solving a technical problem — we’re creating a new model for how sensitive medical research can be conducted responsibly and at scale.
For the families we work with every day, this represents real hope. The barriers that have prevented breakthrough research are being systematically dismantled, and we’re optimistic about the solutions that will emerge from this collaboration.
