AI-Powered Remote Patient Monitoring Platform

From the Department of Computer Science and the Vector Institute for Artificial Intelligence, and proven in collaboration with U of T's affiliated hospitals, this technology platform connects doctors to patients and enables real-time insights from wearables. 

The platform allows for clinical-grade data analysis from wearable sensors (e.g. smartwatches, mobile devices) through proprietary machine learning (ML)-based algorithms that extract clinically relevant data and filter out unreliable sensor data. This platform can provide real-time feedback on patient health, generate more accurate predictions, and enable actionable insights and recommendations to improve care.

Key features:

  • ML-based software to monitor and predict clinically-relevant changes in patient health
  • Patient mobile app for self-reporting
  • Clinician dashboard that seamlessly integrates into workflows to provide clinically-relevant patient health metrics and actionable insights



Advances in AI, telemedicine, and wearable sensor technologies (e.g. smartwatches and mobile phones) represent novel and feasible methods for improved disease management and diagnosis through long-term, continuous patient monitoring, and are reshaping healthcare delivery. Despite these advances, there is still a need for clinically-validated, turnkey solutions for clinicians integrate into their care. 

By combining computer science expertise with clinical experience, the platform helps:

  • Reduce burden on healthcare workers
  • Improve patient outcomes
  • Provide peace of mind to patients
  • Reduce costs

The core IP combines novel filters and a convolutional neural network model to automatically learn what kind of data will make an algorithm produce inaccurate results. While the idea of rejecting certain data is not new, existing filters are manually-developed and rely on assumptions about what causes the algorithm to be unreliable.

In-the-wild continuous sensing on mobile devices has the potential to revolutionize fields such as personalized health care. However, a key problem with current methods is the diverse nature and noise associated with incoming sensor data. Running sensor data processing algorithms on this diverse data can lead to unexpected and poor results because it is difficult for algorithms to anticipate the variety of data that can occur. This tunable system is able to achieve error rates significantly lower than existing approaches, especially for:

  • At‑home monitoring
  • Wearable tech
  • Clinical studies
  • Diagnostics



Related Resources

VPRI Contact

Laurent Moreno

Innovations & Entrepreneurship Manager
Innovations & Partnerships Office (IPO)
(416) 946-0594