The Praesiidium project has successfully concluded, delivering an innovative real-time prediction tool designed to assess the risk of prediabetes in patients. The solution is supported by advanced monitoring technologies, including CheckHealth‘s AI-powered health assistant, VocaVital.
Running from 2023 to February 2026, the project focused on identifying undiagnosed diabetes and enabling early prediction of prediabetes. Its goal was to support timely lifestyle interventions through the use of an AI-based prediction algorithm.
Prediction tool to help prevent prediabetes
Undiagnosed diabetes remains a significant public health challenge, affecting an estimated one in three adults in Europe. Early detection is critical, as prediabetes can often be reversed without medication if addressed in time. This makes accurate and accessible prediction tools essential for preventive care.
The Praesiidium algorithm uses physics-informed machine learning to calculate an individual’s risk of developing prediabetes. It is trained on a combination of clinical data, family history and real-time inputs from wearable sensors that monitor glucose levels, bioimpedance and heart rate.
CheckHealth’s AI-driven health assistant supported patient monitoring
To further support patient monitoring, particularly around dietary habits, CheckHealth developed VocaVital, an AI-driven voice assistant tested in one of the project’s clinical studies. The assistant conducted structured phone interviews to collect nutritional data, which was automatically transcribed, analysed and shared with healthcare professionals as actionable insights.
All collected data was securely managed through the LinkWatch platform, ensuring efficient integration and accessibility for care teams.
Together, these innovations demonstrate how AI and digital health tools can transform early detection and prevention strategies for chronic conditions like diabetes.
CheckHealth is a subsidiary of CNet, providing remote patient monitoring solutions.
