Canada launches ARCHIMEDES, a national platform to break down health data silos

University of Ottawa Heart Institute, March 5, 2026

Canadian health researchers have a data problem. The datasets they need often sit in separate institutional systems that do not easily connect. Getting permission to transfer data between hospitals, universities, and research centers can take months. By the time access is granted, the research question may have moved on.

A new platform called ARCHIMEDES (Advanced Research Collaboration for Health Integration, Medical Exploration, and Data Synthesis) aims to fix this. Launched by the University of Ottawa Heart Institute, the University of Ottawa, and McGill University, the platform provides a centralized, ethically governed environment where researchers can access and combine diverse health datasets, including behavioral data, imaging, genomics, and biospecimens.

How the platform works

ARCHIMEDES uses a two-tier access model. Researchers who contribute data retain control over how it is shared. This is not a system where contributing a dataset means losing authority over it. The model is designed to support both collaboration and compliance with Tri-Council data management policies and bioethical standards.

A data access committee oversees both contribution and sharing. Researchers can combine datasets, explore trends, and run analyses using advanced tools, including AI-driven analytics, predictive modeling, and data visualization. The platform also supports high-performance computing, which is increasingly necessary for the scale of analysis modern health research demands.

Kelly Cobey, co-chair of ARCHIMEDES and a scientist at the Ottawa Heart Institute, framed the problem simply: important health research data too often sits in separate systems that do not connect. The platform provides infrastructure to combine datasets while maintaining public trust.

Mental health data first

The University of Ottawa's Institute of Mental Health Research will be the first user, contributing high-priority mental health data. Additional registered access datasets are planned for upcoming phases.

The choice of mental health as a starting point is notable. Mental health research is particularly fragmented, with clinical, behavioral, and genomic data often stored in entirely separate systems with different formats and access rules. If ARCHIMEDES can bridge those divides for mental health data, the template should extend to other disease domains.

Alan Evans, a professor at McGill University and co-chair, emphasized that the platform was originally developed for brain research but has been generalized for application to all disease domains. The underlying technology builds on world-leading data platforms created for neuroscience.

The problem ARCHIMEDES addresses

Canada's health data landscape is fragmented by design, not malice. Privacy regulations, institutional policies, and provincial jurisdictions all create barriers. These barriers serve important purposes: they protect patient privacy and prevent misuse of sensitive information. But they also slow research.

The result is that datasets with enormous potential value remain siloed. A hospital system's clinical records cannot easily be combined with a university's genomic data, even when doing so could accelerate understanding of how diseases develop and respond to treatment.

ARCHIMEDES attempts to thread the needle between openness and control. By providing a governed environment where data can be accessed without being moved or surrendered, it reduces the friction of collaboration without dismantling the privacy protections that exist for good reason.

Limitations and open questions

The platform's success will depend on adoption. A data-sharing infrastructure is only as valuable as the data it contains, and convincing researchers and institutions to contribute requires sustained trust in the governance framework. Early institutional partnerships are promising, but scaling to a truly national platform will require buy-in from hospitals, provincial health systems, and funding agencies across the country.

There is also the challenge of data harmonization. Health data comes in many formats, and combining datasets from different sources with different collection methods is technically difficult. The platform will need robust tools for data standardization if the combined datasets are to produce reliable results.

ARCHIMEDES is funded through the Brain-Heart Interconnectome Research Program via the Canada First Research Excellence Fund, a partnership between the Ottawa Heart Institute, McGill University, and the University of Ottawa.