A European research project involving Karolinska Institutet will investigate whether artificial intelligence, AI, can help detect heart failure earlier in primary care. The project has also been selected by WHO Europe for presentation at an international conference in Ireland in 2027.
Heart failure is a condition that can be difficult to identify in primary care. As part of the European Faithful project, the clinical study Decision is investigating whether AI-based analysis of electrocardiograms (ECGs) can provide healthcare professionals with better support when assessing patients with suspected heart failure.
The study is being conducted in Sweden and Spain and includes people aged 65 years or older. Participants either have symptoms that raise suspicion of heart failure or at least one risk factor for cardiovascular disease without having any symptoms.

"We want to investigate the value that AI-assisted ECG interpretation can add in primary care, not only in identifying people who may have heart failure but also in the practical healthcare process," says Fernando Seoane , principal investigator at Karolinska Institutet for the Decision/Faithful project, at the Department of Clinical Science, Intervention and Technology.
Participants in the study are randomly assigned to two groups: one in which the physician receives AI support when interpreting the ECG and one in which standard ECG interpretation is used. In the AI group, the Willem™ platform provides an additional risk assessment based on the patient's ECG. The decision on whether to refer a patient to a cardiologist is always made by the physician.
The Swedish part of the study has 1 417 participants, of whom 284 are in the group with symptoms and 1 133 in the group with risk factors.
Assessing benefits for patients and healthcare
The researchers follow the participants for six months. In addition to comparing diagnostic methods, they are studying the experiences of patients and healthcare professionals, the use of healthcare resources and the costs involved. The results will form the basis for a health economic evaluation and provide knowledge about whether AI support can be a cost-effective tool in primary care.
The implementation of the AI support in the Swedish part of the study is supported by Stockholm medical artificial intelligence and learning environments (SMAILE). At Gustavsberg University Primary Care Centre, the support is used in routine clinical practice as part of the study.
"For AI to be used responsibly in healthcare, we need to understand both the clinical effects and how the technology works when it is integrated into the daily work of patients and healthcare professionals," says Fernando Seoane.
WHO highlights experiences from primary care
Faithful has been selected from 153 submissions for presentation at the WHO event Global Expo 2027, which will take place in Ireland in February 2027.
The Faithful consortium is led by Idoven, a health technology company working to advance early detection and precision medicine in cardiovascular disease. The consortium comprises 14 partner organisations in Spain, Sweden, France and Estonia, as well as one global partner.