$1 million awarded to use AI to predict developmental trajectory of kids with increased genetic likelihood of autism and epilepsy
Summary:
The project is funded through the Precision Child Health Partnership Translational Program, made possible by the transformational support of the Azrieli Foundation, and aims to help facilitate earlier interventions for children with neurodevelopmental conditions.
Genetic testing can identify children with a greater genetic likelihood of autism, epilepsy or intellectual disability very early in life, increasingly before symptoms of these conditions manifest. However, even after identification of the genetic variant, it’s difficult for families to understand what to expect for their individual child.
One innovative project, led by joint investigators at The Hospital for Sick Children (SickKids) and CHU Sainte-Justine, is aiming to develop a predictive AI model that can estimate the individual trajectory for each child at higher risk — helping reduce uncertainty for families and facilitate early interventions tailored to each child’s needs.
Drs. Jacob Vorstman (SickKids) and Sébastien Jacquemont (CHU Sainte-Justine), co-leads of the project, were recently awarded $1 million through the Precision Child Health Partnership (PCHP) Translational Program. The PCHP is a partnership between two of Canada’s leading paediatric research hospitals, SickKids and CHU Saint-Justine, aimed to combine strengths and jointly pioneer a new era of Precision Child Health in Canada. It’s funded by a transformational gift from the Azrieli Foundation, which enables initiatives like the PCHP Translational Program to invest in projects where joint teams look at translating discoveries into real-world impact for patients and families.
Their project, called PRAIDICT (Precision Medicine through AI-Driven Investigation of Childhood Neurodevelopmental Trajectories), aims to use advanced AI systems to analyze large amounts of clinical data from both institutions and map patterns that could offer clues that would support each individual child on their neurodevelopmental journey.
With these new insights, the team hopes to address the widening gap between the identification of increased risk thanks to improved access to genetic testing, and the unknowns around how each child’s development may unfold. They also hope to set a new precedent in the Canadian landscape, showcasing what’s possible in connecting data and electronic health records safely to improve care.
If we want to be precise for one child, we first need to learn from a million children
"If we want to be precise for one child, we first need to learn from a million children. There’s a wealth of information available within our health care system, with more emerging every day,” says Vorstman. “This project represents a necessary and important step in transforming that information into actionable, tailored insights for patients and families."
PRAIDICT builds on a foundational project, also led by the duo, that explored how large-scale electronic health record data could be used to better understand complex patterns of health and neurodevelopment in children. That initial project, supported through the PCHP Catalyst Grant Program in 2024, established the infrastructure needed to securely extract and analyze clinical data across institutions and test how they might responsibly use AI.
"Through the initial funding we were able to test new approaches and demonstrate the potential of combining expertise and data across both institutions," says Jacquemont. “Now, with this additional support, we can ensure the benefits reach patients and families at higher risk for neurodevelopmental conditions.”
PCHP continues to strengthen a thriving ecosystem of collaboration between SickKids and CHU Sainte-Justine, unlocking the potential of Precision Child Health and setting an example of how working together can enable more individualized care for children across Canada.