Published in Nature Communications: Novel application of AI to ECG Images

Led by Veer Sangha is our lab’s most recent deep learning application to layout-free automated clinical diagnosis & hidden label detection from ECG images With broad global access to ECG images, this represents an advance for AI in ECG inference.

The study was published in Nature Communications today.

The current standard for AI-ECG is inference on raw signals. However, clinicians in US & globally may not have signal data, but do have printed ECGs. Different layouts and styles also limit AI applications. We developed an approach that addresses this & created a tool ECG DX©.

Check out the coverage of our work from Yale Cardiology below.