Heart failure is the leading cause of hospitalization for people over the age of 65, and studies predict that by 2030, the disease will increase in global prevalence by almost 50%. Currently, more than 64 million people around the world suffer from heart failure. To reduce the burden of disease on both the health system and patients, the Amsterdam UMC is launching a consortium to look for AI-powered solutions, thanks to a grant of around €6 million from Horizon Europe. increase.
Volkert Asselbergs, leader of the consortium and professor of precision medicine at the UMC Amsterdam, said the consortium “has the ambition to develop models that can predict the expected outcome of individual patients based on available data.” We have a purposeful goal that will enable timely and highly personalized care.” Care. “
To achieve this “ambitious” goal, a consortium of partners around the world will combine cardiac imaging, cardiac biomarkers, ECG data, and data from cardiac imaging to develop personalized risk assessments. Develop algorithms to facilitate evaluation. Information from patient records. To build this AI model, the consortium will combine data from approximately 900,000 patients in Europe, South America, and Africa.
Our patient dataset provides an excellent opportunity to really understand all the factors that determine symptom progression and assess risk for individual patients. Furthermore, engaging patients and clinical experts through co-creation workshops from the beginning ensures that all relevant patient and clinical requirements are considered when developing AI models, ensuring comprehensiveness, accountability and transparency. , equity is maximized and goes far beyond quality of care. Technical performance required by the current regulatory framework. ”
Folkert Asselbergs, Professor of Precision Medicine, UMC Amsterdam
Once the AI model is built, it will also translate into an easy-to-use tool for clinicians to share decisions in the booking room. Once the project is complete and the appropriate certifications are obtained, the goal is to play a wider role.
Responsible AI
A key aspect of this project is ensuring that AI models are implemented in a responsible manner. For Asselberg, this is “extremely important” to successfully deploying AI. “Enabling our models requires the ongoing engagement of patients, clinicians and data professionals. This can only happen if there is trust in how the data is used.” he adds Asselbergs.
To foster this trust, the consortium will also develop mechanisms to monitor how the AI performs after being deployed in real-world environments. His five clinical sites involved in this study are located in the Netherlands, Spain, Czech Republic, Peru, and Tanzania, and the diversity of locations not only allows for thorough testing of the model, but also facilitates the collection of diverse patient data. will be
This includes research into social and ethical issues that may arise from the introduction of AI to assess patient risk. “With this project, we want to show how powerful AI can be as a clinical tool, but at the same time consider all the factors that influence the use of AI and do it in a responsible way. We want to run it in , ”he concludes Asselbergs.
sauce:
University of Amsterdam Medical Center
