Artificial Intelligence (AI) in Cardiology: An Electrophysiology Perspective
AI is increasingly enhancing cardiology by improving ECG interpretation, advancing cardiac implantable devices, and addressing challenges, with the potential to revolutionize cardiac care and patient outcomes. Thank you to Stephanie Rollins, RN, CV-BC for this article on AI in cardiology.
Artificial Intelligence (AI) is becoming more prevalent in our daily lives. You may have noticed the auto-correct AI feature on your smartphone when typing or texting. Essentially, AI involves machines or computers using vast amounts of data to solve problems, often more efficiently than humans. In healthcare, this data can be used by scientists and providers to research, diagnose, and treat medical conditions.
While AI may be relatively new, healthcare roles already rely on it. For example, ECG machines analyze measurements down to the millisecond and provide a comprehensive initial interpretation within seconds. However, a cardiologist or physician then assesses and adjusts the ECG findings before submitting the final results.
There are countless ways AI is being used to enhance Cardiology in even more comprehensive ways than before. Three extensively researched areas are ECG for disease prediction, advanced arrhythmia detection, and cardiac device monitoring.
AI-assisted ECG Interpretation
Smartwatches detect atrial fibrillation (AFib) episodes with high accuracy using simplified portable ECGs. Beyond this, the Mayo Clinic is creating a wearable device that identifies low ejection fractions,1 which determines how hard the left ventricle squeezes by measuring the percentage of blood pumped from the left ventricle with each contraction. This feature allows providers to determine asymptomatic left ventricular dysfunction more cost-effectively for patients.2 Technology is already being added to digital stethoscopes to make heart sounds easier to detect and may be able to detect arrhythmias in the near future.
Other AI developments detect subtle details on an ECG that indicate recent AFib episodes, even if the ECG doesn’t show the arrhythmia at the time of the ECG recording.3
Although ECGs already indicate possible diseases such as left ventricular hypertrophy, adding AI formulas showed earlier disease detection than traditional rule-based criteria alone. AI detects ultra-fine ECG variants at times undetectable to human experts and discovers formerly unknown ECG patterns directly related to disease development and progression. One example showed that AI blueprints discovered new ECG indicators for left ventricular hypertrophy criteria. In the same study, researchers found that AI methods estimated the presence of diseases but also their level of progression. The most accurately detected were pulmonary arterial hypertension and hypertrophic cardiomyopathy (HCM). After diagnostics, pulmonary arterial hypertension showed the most robust indicators, including disease progression, confirmed with simultaneous studies.4
Cardiac Implanted Electronic Devices
Cardiac Implanted Electronic Devices (CIED) such as pacemakers and defibrillators already have decades of AI use with ever-expanding technology upgrades. These devices have numerous settings to evaluate a person’s heart rate and rhythm down to the millisecond and respond quickly with preset therapies as needed. These therapies provide a tiny dose of electricity to stimulate the heart to beat faster or deliver defibrillator therapy during a sudden rapid heartbeat or cardiac arrest.
Over the last several years, leadless pacemakers have been developed as an alternative to traditional pacemakers. Leadless pacemakers were created as single-chamber devices for right ventricular pacing. Medtronic later developed Micra AV, a leadless pacemaker that senses atrial activity. This leadless pacemaker relies on mechanical atrial contractions that signal the device to deliver right ventricular pacing in response.5,6 Most recently, device company Abbott released a leadless pacemaker for the right atrium, allowing both atrial and ventricular leadless pacing.7
Another recent advancement in cardiac remote monitoring involves Implanted Cardiac Monitors (ICMs). ICMs are inserted just under the skin to detect heart arrhythmias over a period of a few years and are frequently used to detect AFib, pauses in heartbeats, and other arrhythmias. Device company Medtronic’s ICMs now use AI advancements that detect AFib and pause occurrences at a 99-100% accuracy rate and significantly reduce false positives.8 In addition to these detections, device company Biotronik’s BioMonitor IV with Smart ECG also uses AI to distinguish between premature atrial and ventricular contractions and reduces false positives throughout all significant arrhythmias.9 These improvements assist device clinics by decreasing the time spent reviewing false positive alerts, allowing more substantial time with direct patient care.
AI is not only improving the accuracy of CIEDs, but it’s also changing the lives of patients with heart failure. Heart failure-specific CIEDs monitor fluid levels in the thoracic cavity, discern fluid-sensitive changes in cardiac vibrations during specific moments of the cardiac cycle, variations in a person’s activity level, respiratory rate, and more. The healthcare team then uses this data to collaborate with physicians to decrease the frequency of heart failure exacerbation. Often, these tools show changes before patients have significant symptoms.10
Challenges Incorporating AI in Cardiology
Ensuring that AI is free from biases is crucial for achieving optimal treatment outcomes for everyone. It’s important to gather data representing a large and diverse population to avoid any potential biases arising from the data pool that AI learns from. Additionally, there is a concern that AI may decrease the direct relationship between clients and providers. However, informing clients about any technological changes can increase their peace of mind and ultimately strengthen the client-provider relationship.11 It is essential to strike a balance between leveraging the benefits of AI and maintaining the human touch in healthcare settings.
Benefits for Everyday Providers
In conclusion, AI technology has the potential to greatly improve the field of cardiology by providing faster and more comprehensive data, personalized treatment options, and greater patient involvement in decision-making. While some may be concerned about technology replacing healthcare providers, it can free up more time for direct patient care. Furthermore, community clinics can benefit from this technology by providing faster, more accessible, affordable cardiac care. Overall, AI has the power to revolutionize the way we approach cardiac care and ultimately improve patient outcomes.
- Atrial Fibrillation Patient Tools and Handouts
- Atrial Fibrillation Resources for Providers
- How Cardiac Rehab Can Improve Functional Capacity in Patients with LVAD
- Friedman, P. A., M.D., Kapa, S., M.D., Lopez-Jimenez, F., M.D., M.B.A., & Noseworthy, P. A., M.D. (2023, September 12). Artificial Intelligence (AI) in Cardiovascular Medicine. The Mayo Clinic Cardiovascular Medicine.
- Attia, Z. I., Kapa, S., Lopez-Jimenez, F., McKie, P. M., Ladewig, D. J., Satam, G., Pellikka, P. A., Enriquez-Sarano, M., Noseworthy, P. A., Munger, T. M., Asirvatham, S. J., Scott, C. G., Carter, R. E., & Friedman, P. A. (2019). Screening for cardiac contractile dysfunction using an artificial intelligence-enabled electrocardiogram. Nature medicine, 25(1), 70–74.
- Noseworthy, P. A., M.D., et al. (2019, December 30). Artificial intelligence-guided screening for atrial fibrillation using electrocardiogram during sinus rhythm: A prospective non-randomised interventional trial. National Library of Medicine National Center for Biotechnology Information. Retrieved September 7, 2023.
- Tison, G. H., Zhang, J., Delling, F. N., & Deo, R. C. (2019, September 5). Automated and Interpretable Patient ECG Profiles for Disease Detection, Tracking, and Discovery.
- Bhatia, N. K., & Merchant, F. M. (2021). Atrio‐ventricular synchronous pacing with a single chamber leadless pacemaker: Programming and trouble shooting for common clinical scenarios. Journal of Cardiovascular Electrophysiology, 32(2), 533-539.
- Medtronic (n.d.). Micra leadless pacemakers are the world’s smallest pacemakers for bradyarrhythmia management. Micra™ AV2 and Micra™ VR2 leadless pacemakers.
- Abbott (n.d.). AVEIR DR Dual Chamber Leadless Pacemaker System.
- Medtronic (n.d.). AccuRhythm™ AI algorithms.
- Biotronic (n.d.). BIOMONITOR IV with SmartECG. Biotronic Excellence for Life.
- Boston Scientific (n.d.). HeartLogic Clinical Data. Boston Scientific Advancing Science For Life.
- Nick, S. (2019, December 30). AI Won’t Replace Your Doctors but it could make them much better. Discovery’s Edge, Mayo Clinic’s Research Magazine. Retrieved September 7, 2023, from