|
Project Topic
|
PerCard explores the value of combining integrated heterogeneous data sources with AI/ML to increase the validity of risk assessment for cardiovascular disease in different populations. The consortium develops specialised models for populations and explicitly mitigates the issue of gender-bias in existing risk assessment methods. Emphasis is put on methods that are easily transferable from research to practice and accessible and affordable for the wide population. PerCard develops and uses new data analysis methods (AI, ML, signal processing, including ensemble classifiers and Bayesian neural networks) to develop more powerful risk assessment methods that are accurate, robust, and explainable. The project provides advancements especially in the field of explainable AI and AI for time-series analysis. PerCard delivers an integrated decision support system for understandable personalised risk/prognosis assessment of morbidity and mortality in cardiovascular settings and tests this robustly in retrospective and prospective settings. End-users are involved throughout the entire research, development and validation process. ELSA considerations, including gender, accessibility to all, ethics of AI and decision support are an inherent part of the project. To capitalise on earlier investments and speed up development, PerCard re-uses large retrospective databases to start with (Finland, Italy). PerCard uses robust validation methods to guarantee generalisability, combining independent test sets as well as prospective test data collection as truly independent reference. To succeed, PerCard brings together an international interdisciplinary consortium of top-level partners: Tampere University (TUNI, co-ordinator, data analysis and link to Finnish clinical research), Polytechnic University of Milan (POLIMI, data analysis), Centro Cardiologico Monzino (CCM, clinical research), and Protestant University of Applied Sciences Ludwigsburg, Germany (PUL, ELSA considerations).
|
