The new study, supported by the EU-funded euCanSHare and HealthyCloud projects, has revealed that a model combining ECG features and cardiac imaging-derived radiomics data improves the detection of Atrial fibrillation (AF, i.e., a heart condition characterised by an irregular and often abnormally fast heart rhythm) in women.

The main clinical tool for diagnosing AF, that can cause problems such as dizziness, shortness of breath and tiredness, and increases the risk of stroke and heart failure, is the Electrocardiogram (ECG), widely used to spot abnormalities in heart rhythms and waveforms.

However, an ECG recorded at a single time point may not detect individuals with paroxysmal AF. In this study, machine learning models have been developed for discrimination of prevalent AF using a combination of image-derived radiomics phenotypes and ECG features.

To read more: https://www.nature.com/articles/s41598-022-21663-w