Abstract:Timely and accurate assessment of electrocardiograms (ECGs) is crucial for diagnosing, triaging, and clinically managing patients. Current workflows rely on a computerized ECG interpretation using rule-based tools built into the ECG signal acquisition systems with limited accuracy and flexibility. In low-resource settings, specialists must review every single ECG for such decisions, as these computerized interpretations are not available. Additionally, high-quality interpretations are even more essential in such low-resource settings as there is a higher burden of accuracy for automated reads when access to experts is limited. Artificial Intelligence (AI)-based systems have the prospect of greater accuracy yet are frequently limited to a narrow range of conditions and do not replicate the full diagnostic range. Moreover, these models often require raw signal data, which are unavailable to physicians and necessitate costly technical integrations that are currently limited. To overcome these challenges, we developed and validated a format-independent vision encoder-decoder model – ECG-GPT – that can generate free-text, expert-level diagnosis statements directly from ECG images. The model shows robust performance, validated on 2.6 million ECGs across 6 geographically distinct health settings: (1) 2 large and diverse US health systems-Yale-New Haven and Mount Sinai Health Systems, (2) a consecutive ECG dataset from a central ECG repository from Minas Gerais, Brazil, (3) the prospective cohort study, UK Biobank, (4) a Germany-based, publicly available repository, PTB-XL, and (5) a community hospital in Missouri. The model demonstrated consistently high performance (AUROC≥0.81) across a wide range of rhythm and conduction disorders. This can be easily accessed via a web-based application capable of receiving ECG images and represents a scalable and accessible strategy for generating accurate, expert-level reports from images of ECGs, enabling accurate triage of patients globally, especially in low-resource settings.

A community effort to assess and improve computerized interpretation of 12-lead resting electrocardiogram

High-Performance Personalized Heartbeat Classification Model for Long-Term ECG Signal

DeepECG: Image-based Electrocardiogram Interpretation with Deep Convolutional Neural Networks

LabelECG: A Web-based Tool for Distributed Electrocardiogram Annotation

ECG-CL: A Comprehensive Electrocardiogram Interpretation Method Based on Continual Learning

Cardiologist-level interpretable knowledge-fused deep neural network for automatic arrhythmia diagnosis

A 12-lead electrocardiogram database for arrhythmia research covering more than 10,000 patients

Identification of 27 abnormalities from multi-lead ECG signals: an ensembled SE_ResNet framework with Sign Loss function

Interpretable Deep Learning for Automatic Diagnosis of 12-lead Electrocardiogram

Multi-label Classification of Abnormalities in 12-Lead ECG Using Deep Learning.

Automatic Detection of ECG Abnormalities by using an Ensemble of Deep Residual Networks with Attention

Assisting Schizophrenia Diagnosis Using Clinical Electroencephalography and Interpretable Graph Neural Networks: a Real-World and Cross-Site Study

Practical Lessons on 12-Lead ECG Classification: Meta-Analysis of Methods From PhysioNet/Computing in Cardiology Challenge 2020

ECG signal classification based on deep CNN and BiLSTM

CRT-Net: A Generalized and Scalable Framework for the Computer-Aided Diagnosis of Electrocardiogram Signals

Large-scale cross-modality pretrained model enhances cardiovascular state estimation and cardiomyopathy detection from electrocardiograms: An AI system development and multi-center validation study

An artificial intelligence–enabled ECG algorithm for comprehensive ECG interpretation: Can it pass the ‘Turing test’?

Artificial Intelligence System for Detection and Screening of Cardiac Abnormalities using Electrocardiogram Images

Beat-Level Interpretation of Intra-Patient Paradigm Based on Object Detection

ECG-Image-Database: A Dataset of ECG Images with Real-World Imaging and Scanning Artifacts; A Foundation for Computerized ECG Image Digitization and Analysis

Automated Diagnostic Reports from Images of Electrocardiograms at the Point-of-Care