早期妊娠心电图特征研究与人工智能诊断

Research of ECG Characteristics and Artificial Intelligent Diagnosis for Early Pregnancy

摘要: 目的：研究青年女性早期妊娠心电图的特征性改变，为早期妊娠的女性提供一种新的早期筛查手段，提高早期妊娠的诊断和干预率。方法：152例早期妊娠的青年女性以及100例未妊娠的青年女性，通过医学SPSS软件对比两组人群的心电图差异，同时运用深度学习方法对心电图进行智能诊断，开发诊断软件。结果：早期妊娠女性的心电图出现窦性心动过速、窦性心律不齐、短PR间期、ST段压低、T波低平倒置、电轴左偏、胸导联低电压、逆钟向转位较未妊娠女性具有显著的统计学差异(P < 0.05)。智能诊断软件预测准确率达到90%，精确率100%，召回率83.33%。结论：青年女性早期妊娠心电图均为生理性变异，基于人工智能心电图诊断早期妊娠准确率、精确度均较高。

Abstract: Objectives: This paper detects the changes of electrocardiogram (ECG) characteristics of young women in early pregnancy, providing a new screening method for women in early pregnancy and improving the diagnosis and intervention rate of early pregnancy. Methods: A total of 152 young women with early pregnancy and 100 young women without pregnancy were included. The ECG dif-ferences between the two groups were compared by SPSS software, and deep learning method is used for intelligent diagnosis of ECG. A diagnostic software was developed. Results: The ECG charac-teristics of early pregnant women were sinus tachycardia, sinus arrhythmia, short PR interval, ST segment depression, T wave inversion, left axis deviation, chest lead low voltage and inverse clock transposition, which had significant statistical differences compared with non-pregnant women (P < 0.05). The accuracy rate, precision rate and recall rate of intelligent diagnostic software reach 90%, 100%, and 83.33% respectively. Conclusions: The ECG characteristics of young women in early pregnancy are physiological variations. The accuracy and precision of early pregnancy diagnosis based on artificial intelligence are high.