MEIT: Multimodal Electrocardiogram Instruction Tuning on Large Language Models for Report Generation
This work addresses the need for efficient and versatile ECG report generation in clinical settings, representing a novel application of LLMs to a specific medical domain.
The authors tackled the problem of automating ECG report generation, which is time-consuming and requires clinical expertise, by proposing the MEIT framework that uses multimodal instruction tuning with LLMs, achieving superior performance in quality report generation, zero-shot capabilities, and alignment with human expert evaluation across over 800,000 ECG reports.
Electrocardiogram (ECG) is the primary non-invasive diagnostic tool for monitoring cardiac conditions and is crucial in assisting clinicians. Recent studies have concentrated on classifying cardiac conditions using ECG data but have overlooked ECG report generation, which is time-consuming and requires clinical expertise. To automate ECG report generation and ensure its versatility, we propose the Multimodal ECG Instruction Tuning (MEIT) framework, the first attempt to tackle ECG report generation with LLMs and multimodal instructions. To facilitate future research, we establish a benchmark to evaluate MEIT with various LLMs backbones across two large-scale ECG datasets. Our approach uniquely aligns the representations of the ECG signal and the report, and we conduct extensive experiments to benchmark MEIT with nine open-source LLMs using more than 800,000 ECG reports. MEIT's results underscore the superior performance of instruction-tuned LLMs, showcasing their proficiency in quality report generation, zero-shot capabilities, resilience to signal perturbation, and alignment with human expert evaluation. These findings emphasize the efficacy of MEIT and its potential for real-world clinical application.