首页|肢体导联组合对心电轴估计精度影响的研究

肢体导联组合对心电轴估计精度影响的研究

扫码查看
目的 心电肢体导联共有6通道,任意2导组合均可用于心电轴的估计.本研究比较了15种肢体导联组合的心电轴估计精度.方法 使用生理数据网的12导心电图公开数据库(每条记录时长10 s,采样率500 Hz,静息状态),抽取年龄≥18岁且标注为单一窦性节律类型(心电轴正常)的21 306条心电记录,包括6 153条窦性节律记录、10 916条窦性心动过缓记录、3 466条窦性心动过速记录、771条窦性不规则记录;抽取年龄≥18岁的2 323条心电轴偏移记录,包括1 526条心电轴左偏记录、797条心电轴右偏记录.分别通过{Ⅰ,Ⅱ}、{Ⅰ,Ⅲ}、{Ⅰ,aVR}、{Ⅰ,aVL}、{Ⅰ,aVF}、{Ⅱ,Ⅲ}、{Ⅱ,aVR}、{Ⅱ,aVL}、{Ⅱ,aVF}、{Ⅲ,aVR}、{Ⅲ,aVL}、{Ⅲ,aVF}、{aVR,aVL}、{aVR,aVF}、{aVL,aVF}共15种"导联对"的QRS波群幅度(或面积)的正、负代数和来估计心电轴.结果 基于幅度方法,{Ⅰ,Ⅱ}、{Ⅱ,aVL}对心电轴正常、心电轴左偏、心电轴右偏的三分类精度依次为93.56%与93.50%,优于传统{Ⅰ,aVF}的识别精度(92.93%).基于面积方法,{Ⅲ,aVR}、{Ⅰ,aVR}、{Ⅰ,Ⅱ}、{aVR,aVF}、{Ⅱ,aVL}、{Ⅱ,Ⅲ}对心电轴正常、心电轴左偏、心电轴右偏的三分类精度依次为92.66%、92.53%、92.29%、92.19%、92.10%与91.91%,优于传统的{Ⅰ,aVF}识别精度(91.82%).结论 幅度方法的精度优于面积方法;无论幅度方法还是面积方法,心电轴自动估计时"导联对"{Ⅰ,Ⅱ}与{Ⅱ,aVL}的精度均高于传统"导联对"{Ⅰ,aVF}.
Influence of limb-lead combination on the accuracy of the estimation of cardiac electrical axis
Objective There are 6 leads of limb lead ECG,and the cardiac electrical axis can be estimated by any combination of two leads.In this paper,the estimation accuracy of all 15 pairs of limb-lead combinations was compared.Methods Using the open database of 12-lead electrocardiograms(at a sampling frequency of 500 Hz with duration of 10 seconds during resting state)from PhysioNet,totally 21 306 ECG records were extracted with age≥18 years which labeled as single sinus type(axis normal),including 6 153 records with Sinus Rhythm,10 916 records with Sinus Bradycardia,3 466 records with Sinus Tachycardia,and 771 records with Sinus Irregularity.Moreover,totally 2 323 axis deflection recordings with age≥18 years were extracted,including 1 526 records with Axis left shift,and 797 records with Axis right shift.Cardiac electrical axis was estimated with the net amplitude(or area)of QRS complex(algebraic sum of positive and negative amplitude or area)by any pair of leads from{Ⅰ,Ⅱ},{Ⅰ,Ⅲ},{Ⅰ,aVR},{Ⅰ,aVL},{Ⅰ,aVF},{Ⅱ,Ⅲ},{Ⅱ,aVR},{Ⅱ,aVL},{Ⅱ,aVF},{Ⅲ,aVR},{Ⅲ,aVL},{Ⅲ,aVF},{aVR,aVL},{aVR,aVF},{aVL,aVF},respectively.Results For the amplitude-based method,the recognition accuracy for the normal,left and right axes from{Ⅰ,Ⅱ}and{Ⅱ,aVL}is 93.56%and 93.50%,respectively,which is better than that of the traditional classical method{I,aVF}(92.93%).For the area-based method,the recognition accuracy from{Ⅲ,aVR},{Ⅰ,aVR},{Ⅰ,Ⅱ},{aVR,aVF},{Ⅱ,aVL}and{Ⅱ Ⅲ}is 92.66%,92.53%,92.29%,92.19%,92.10%and 91.91%,respectively,which is better than the traditional classical method{Ⅰ,aVF}(91.82%).Conclusion The accuracy of amplitude-based method is higher than that of area-based method.Lead pair{Ⅰ,Ⅱ}and{Ⅱ,aVL}have higher accuracy than traditional classical{Ⅰ,aVF}in automatic estimation of cardiac electrical axis for both amplitude and area method.

ECGcardiac electrical axisQRS axis,axis deflectionaxis left shiftaxis right shiftlimb leads

李延军、刘伟波、宫国强、张琳、宋晋忠、杨宜婧

展开 >

中国航天员科研训练中心,北京 100094

心电 心电轴 QRS轴 心电轴偏移 心电轴左偏 心电轴右偏 肢体导联

中国载人航天生物医学工程预先研究项目

2024

航天医学与医学工程
中国航天员科研训练中心

航天医学与医学工程

影响因子:0.392
ISSN:1002-0837
年,卷(期):2024.35(3)
  • 4