多模式大语言模型与专用心电图AI的比较诊断性能,用于从心电图图像检测心肌梗塞:比较研究
Haemin Lee1,2, Sooyoung Yoo3, Joonghee Kim1,2
1Department of Emergency Medicine, Seoul National University Bundang Hospital, Seongnam-si, Gyeonggi-do, Republic of Korea.
JMIR AI
|September 17, 2025
概括
与专门的AI工具相比,一般的大型语言模型,如ChatGPT和Gemini在ECG诊断心肌梗塞 (MI) 的表现不佳. 特定领域的人工智能对于准确的临床解释至关重要.
科学领域:
- 心脏病学和人工智能的人工智能
- 医学成像分析 医学成像分析
- 医疗保健中的机器学习
背景情况:
- 准确的心电图 (ECG) 解释对于紧急心肌梗塞 (MI) 诊断至关重要.
- 多模式大语言模型 (LLM) 显示出临床解释的潜力,但它们依赖波形数据与文本线索的依赖性尚不清楚.
- 需要直接比较一般的LLM和专门的ECGAI工具.
研究的目的:
- 评估ChatGPT和双子座的诊断性能,以从心电图像中检测心脏病发作.
- 为了比较这些一般的LLM与专门的AI工具ECG Buddy的性能.
主要方法:
- 来自巴基斯坦的公开可用的12个心电图数据集 (239个MI阳性,689个MI阴性病例) 的追溯分析.
- 聊天GPT (GPT-4o) 和双子座 (2.5 Pro) 被查询MI的信心.
- ECG Buddy分析了对ST升高MI,急性冠状动脉综合征和心肌损伤生物标志物的ECG图像.
主要成果:
- 电脑心电图Buddy获得了高精度 (96.98%),AUC (98.8%),灵敏度 (96.65%) 和特异性 (97.10%).
- 聊天GPT表现较差:准确率为65.95%,AUC为57.34%,灵敏度为36.40%,特异性为76.2%.
- 双子 2.5 Pro表现不佳:准确率为29.63%,AUC为51.63%,灵敏度为97.07%,特异性为6.24%.
- 对诊断的LLM解释往往是不准确的;ECG Buddy显著优于两种LLM (P<.001).
结论:
- 像ChatGPT和Gemini这样的一般LLM与基于ECG的MI诊断的ECGBuddy等专业AI工具相比表现明显差.
- 特定领域的AI模型对于在临床环境中实现可靠和强大的诊断准确性至关重要.
- 虽然LLM可能会随着进一步的培训而得到改善,但专门的AI目前仍然对ECG准确检测MI至关重要.
相关概念视频
01:25Correlation between ECG and Cardiac Cycle
11.7K
The electrical signals recorded on an electrocardiogram (ECG) occur before the mechanical processes of contraction and relaxation during the cardiac cycle.
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...
11.7K
01:29Electrocardiogram
5.4K
An electrocardiogram (ECG or EKG) is a critical diagnostic tool that records the electrical signals produced by the heart during each heartbeat. This recording is achieved through electrodes placed strategically on the arms, legs, and chest. The electrocardiograph amplifies these signals and produces 12 distinct tracings, offering a comprehensive understanding of the heart's electrical activity.
Three major waveforms are present in a typical ECG recording: the P wave, the QRS complex, and...
Three major waveforms are present in a typical ECG recording: the P wave, the QRS complex, and...
5.4K
01:17Imaging Studies for Cardiovascular System I:Echocardiography
731
Cardiac imaging studies encompass a wide range of noninvasive and minimally invasive techniques designed to visualize the heart's structure and function in detail. One such technique is echocardiography, which uses high-frequency ultrasound waves to produce detailed images of the heart, known as echocardiograms.
Indications: Echocardiography is utilized to diagnose heart failure, valve disorders, and myocardial infarction. It also assesses cardiac structures' size, shape, and motion,...
Indications: Echocardiography is utilized to diagnose heart failure, valve disorders, and myocardial infarction. It also assesses cardiac structures' size, shape, and motion,...
731