The future of artificial intelligence in cardiology and cardiac surgery

The future of artificial intelligence in cardiology and cardiac surgery holds forth the promise of attaining precision cardiovascular medicine and cardiovascular population health. This ideal vision depends, however, on the synergy between cardiologists, cardiac surgeons, cardiac health practitioners, other health professionals, and data scientists. New cardiological and cardiac surgery knowledge will come from unsupervised or self-supervised learning that mitigates the burden of clinicians involved in cardiovascular research projects. Deep learning must occur within the context of a cognitive architecture for future dividends in cardiovascular medicine. Artificial intelligence in imaging will advance exponentially as novel imaging technologies such as few-shot learning, generative adversarial networks, and transfer learning are adopted. In addition, artificial intelligence will converge with emerging technologies such as extended reality to enable medical education and clinical training in cardiology and cardiac surgery. The advent of digital twin technology will facilitate a higher quality of patient care and clinical research that will accelerate with federated and swarm learning. Artificial intelligence and its adoption may be slowed, however, by a lack of adequate access to cardiac healthcare data, specifically a lack of population disease data from cardiac images, biomarkers, and other phenotypic expressions of cardiovascular disease. Artificial intelligence will also face the daunting challenges of clinician education and adoption and legal, regulatory, ethical, and financial issues. A cohort of dually trained and educated cardiologists or cardiac surgeon-data scientists can accelerate the aforementioned dimensions as essential liaisons for both domains. The best dividend from future artificial intelligence in cardiology and cardiac surgery is, perhaps, a reclaiming of humanity in cardiovascular medicine.