Time to capitalise on artificial intelligence in cardiac electrophysiology

The complexity and heterogeneity of data available for patients with cardiovascular disease present an ideal opportunity for artificial intelligence to enhance patient care. For patients with or under investigation for arrhythmias, data are repeatedly and frequently obtained from electrocardiograms, wearable devices, implantable cardiac devices, and during catheter-based electrophysiological procedures [1]. However, whilst artificial intelligence-enhanced electrocardiogram analysis has attracted attention [2], the application of artificial intelligence to other data sources within electrophysiology appears comparatively underdeveloped.
We sought to quantify and compare artificial intelligence research outputs between cardiac and non-cardiac healthcare specialities. We further sought to quantify research outputs using commonly available data types within electrophysiology. We performed a literature search on PubMed using medical subject headings (MeSH) for the term “artificial intelligence” followed by either the name of the healthcare speciality, the name of the cardiac sub-speciality, or the type of electrophysiology data resource (Fig. 1). We subsequently identified the numbers of published articles from inception to 17th March 2024.
The results of our searches are shown in Fig. 1. Oncology (n = 32322), radiology (n = 23613), and pathology (n=22299) had the highest numbers of published articles related to artificial intelligence. Cardiology (n = 4177) was positioned in twelfth place. Amongst cardiac sub-specialities, cardiac electrophysiology (n=193) had the lowest number of artificial intelligence-related research outputs. Within electrophysiology, the application of artificial intelligence to the electrocardiogram was most widely studied amongst the commonly available data types. The search “artificial intelligence electroanatomic” yielded only 32 results.