JMIR Publications

ABSTRACT

Background:

Recent progress has demonstrated the potential of deep learning models in analyzing ECG pathologies. However, this method is intricate, expensive to develop, and designed for specific purposes. Large language models show promise in medical image interpretation, yet their effectiveness in ECG analysis remains understudied. GPT-4o, a multimodal AI model, capable of processing images and text without task-specific training, may offer an accessible alternative.

Objective:

This study evaluates GPT-4o's effectiveness in interpreting 12-lead ECGs, assessing classification accuracy, and exploring methods to enhance its performance.

Methods:

Six common ECG diagnoses were evaluated: Normal ECG, STEMI, AF, RBBB, LBBB, and paced rhythm, with 30 Normal ECGs and 10 of each abnormal pattern, totaling 80 cases (n=80). De-identified ECGs were analyzed using OpenAI’s GPT-4o. Our study employed both zero-shot and few-shot learning methodologies to investigate three main scenarios: (1) ECG image recognition, (2) binary classification of normal versus abnormal ECGs, and (3) multiclass classification into six categories.

Results:

The model excelled in recognizing ECG images, achieving an accuracy of 100%. In the classification of normal/abnormal ECG cases, the Few-Shot learning approach improved GPT-4o’s accuracy by 27%, reaching 80%. However, multiclass classification for a specific pathology remained limited, achieving only 41% accuracy.

Conclusions:

GPT-4o effectively differentiates normal from abnormal ECGs, suggesting its potential as an accessible AI-assisted triage tool. Although limited in diagnosing specific cardiac conditions, GPT-4o’s capability to interpret ECG images without specialized training highlights its potential for preliminary ECG interpretation in clinical and remote settings.