Image-Based ECG Algorithm Uses AI to Diagnose Key Cardiac Disorders

As mobile technology improves, patients increasingly have access to electrocardiogram (ECG) images, which raises new questions about how to incorporate these devices in patient care. Machine learning (ML) approaches, specifically those using deep learning, have transformed automated diagnostic decision-making. For ECGs, they have led to the development of tools that allow clinicians to find hidden or complex patterns. However, deep learning tools use signal-based models, which have not been optimized for remote health care settings. Researchers have now developed an artificial intelligence (AI)-based model for clinical diagnosis that can use ECG images, regardless of format or layout, to diagnose multiple heart rhythm and conduction disorders.

The Yale Cardiovascular Data Science (CarDS) Lab (New Haven, CT, USA) has developed ECG Dx, a novel multi-label automated diagnosis model from ECG images designed to make AI-based ECG interpretation accessible in remote settings. The researchers who analyze multi-modal inputs from electronic health records to design potential solutions hope the new technology provides an improved method to diagnose key cardiac disorders. The model is based on data collected from more than two million ECGs from 1,506,112 patients who received care in Brazil from 2010-2017. One in six patients was diagnosed with rhythm disorders. The tool was independently validated through multiple international data sources, with high accuracy for clinical diagnosis from ECGs. There are a number of clinical and technical challenges when using AI-based applications.

“Current AI tools rely on raw electrocardiographic signals instead of stored images, which are far more common as ECGs are often printed and scanned as images. Also, many AI-based diagnostic tools are designed for individual clinical disorders, and therefore, may have limited utility in a clinical setting where multiple ECG abnormalities co-occur,” said Rohan Khera MD, MS, assistant professor in cardiovascular medicine, who led the team of researchers. “A key advance is that the technology is designed to be smart - it is not dependent on specific ECG layouts and can adapt to existing variations and new layouts. In that respect, it can perform like expert human readers, identifying multiple clinical diagnoses across different formats of printed ECGs that vary across hospitals and countries.”

Related Links:
CarDS Lab