AI Technology Boosts ECG Capabilities for Early Heart Disease Diagnosis

Cardiovascular diseases often remain undetected until a critical event like a heart attack or stroke occurs. Early identification is key to improving outcomes, but the absence of clear symptoms complicates this process. Advances in artificial intelligence (AI) technology are now enhancing the capabilities of the electrocardiogram (ECG), a century-old diagnostic tool, potentially enabling earlier detection and monitoring of heart diseases. Researchers at Mayo Clinic (Rochester, MN, USA) have pioneered the development of ECG-AI algorithms. These algorithms are currently used in research settings to assess the likelihood of various heart conditions, such as atrial fibrillation, amyloidosis, aortic stenosis, low ejection fraction, and hypertrophic cardiomyopathy (HCM). They also use AI to estimate a patient's biological age from both traditional 12-lead ECGs and single-lead ECGs obtained from smartwatches and other portable devices.

The ECG-AI is particularly effective in detecting low ejection fraction, a condition where the heart weakens and pumps less blood. Often, the symptoms of this condition are overlooked or attributed to other causes, like pregnancy, leading to a late diagnosis. ECG-AI can also identify peripartum cardiomyopathy, a type of heart muscle weakness that occurs during or after pregnancy. Similarly, it can help in the early detection of amyloidosis, a rare disease characterized by the buildup of misfolded proteins in organs, which can lead to heart failure.

HCM, a common genetic heart disease, is another condition where ECG-AI proves beneficial. HCM often goes unnoticed as it may not be evident in basic tests like a traditional ECG. However, ECG-AI can identify HCM by detecting patterns that might be missed even by expert clinicians. The potential of ECG-AI to spot these subtle indications early can be crucial in improving patient outcomes and treatment strategies for various heart conditions.

Related Links:
Mayo Clinic