The project shall develop, integrate, train and clinically validate an AI-based system for outcome prediction. The project focus is on predicting atrial fibrillation, but the system can be extended to include prediction of other cardiac diseases.
The AI-based system involves several machine learning models that are developed and trained on individual patient profiles with known outcomes. A patient profile gathers rich content from the patient’s ECGs recorded with Coala Heart Monitor and analysed using the Cardiolund ECG Parser.
The software system shall as part of the project be integrated into the Coala Life medical system, and performance shall be evaluated in a clinical study involving patients from Region Gävleborg.
The project will enable an individualized and prevention-oriented heart screening for users in their home environment. The user will get an early warning when hearth rhythm or beat deviations are interpreted by the outcome prediction model to indicate an elevated risk of cardiac disease. An early warning provides the opportunity for an early reaction by the individual, the warning can motivate the individual to engage in a preventive effort of reducing risk factors. The ultimate goal is to improve the heart health of the individual, and the public health in general.