基于智能腕表的单导联心电图算法识别窦性心动过速及快心室率心房颤动的准确性  

Accuracy of single-lead electrocardiogram algorithm based on intelligent wristwatches in identifying sinus tachycardia and atrial fibrillation with rapid ventricular rate

在线阅读下载全文

作  者:王泓 王浩[2] 张慧[1] 金志赓 邰美慧 郭豫涛[1] Wang Hong;Wang Hao;Zhang Hui;Jin Zhigeng;Tai Meihui;Guo Yutao(Department of Cardiopulmonary Vascular and Thrombotic Diseases,Sixth Medical Department,Chinese PLA General Hospital,Beijing 100048,China;Department of Cardiology,Sencond Medical Department,Chinese PLA General Hospital,Beijing 100853,China;Graduate School,Chinese PLA General Hospital,Beijing 100853,China)

机构地区:[1]解放军总医院第六医学中心肺血管与血栓性疾病科,北京100048 [2]解放军总医院第二医学中心心血管内科,北京100853 [3]解放军总医院研究生院,北京100853

出  处:《中华健康管理学杂志》2023年第11期816-820,共5页Chinese Journal of Health Management

基  金:国家自然科学基金(82170309)。

摘  要:目的分析基于智能腕表的单导联心电图(iECG)算法识别窦性心动过速及快心室率心房颤动(房颤)的准确性。方法本研究为非随机对照试验,于2020年12月15日至2022年5月30日在解放军总医院招募642例≥18岁窦性心动过速(心率111~145次/min)或快心室率房颤(心率110~150次/min)患者为受试者,使其左手腕佩戴华为Watch GT2 Pro智能腕表,将放松状态下腕表检测的生理信号作为实测数据,采用华为公司基于智能腕表的iECG算法进行识别,同时进行12导联心电图(12L-ECG)检查,并由2名心内科医师进行判读作为金标准。根据检测结果,排除不符合纳排标准的受试者3例,最终入组639例为研究对象。采用召回率、精确率、多分类的综合准确率宏观F1值评价该算法识别窦性心动过速及快心室率房颤的准确性。结果纳入分析的639例受试者中,男性469例,女性170例,窦性心动过速389例,快心室率房颤250例,年龄(46.53±13.32)岁。iECG算法识别窦性心动过速的召回率为98.7%,精确率为99.2%,F1值为99.0%;识别快心室率房颤的召回率为98.8%,精确率为98.0%,F1值为98.4%;识别窦性心动过速及快心室率房颤的二分类宏观F1值为98.7%。基于智能腕表的iECG及对应的12L-ECG波形表现出良好的一致性。结论基于智能腕表的iECG算法可有效识别窦性心动过速及快心室率房颤,并表现出良好的准确性。Objective To analyze the accuracy of a single-lead electrocardiogram(iECG)algorithm based on intelligent wristwatch in identifying sinus tachycardia and atrial fibrillation(AF)with rapid ventricular rate.Methods In this non-randomized control trial,642 patients aged≥18 years were enrolled in the General Hospital of Chinese PLA between December 15,2020 and May 30,2022,with sinus tachycardia or rapid ventricular rate of AF(ranging from 111 to 145 beats/min for sinus tachycardia,from 110 to 150 beats/min for rapid ventricular rate of AF,respectively).The patients wore Huawei Watch GT2 Pro smartwatches on their left wrists,and the physiological signals detected by the smartwatches in a relaxed state were used as the measured data.The iECG algorithm developed by Huawei was used for identification.Simultaneously,12-lead electrocardiograms(12L-ECG)were performed,and two cardiologists served as the gold standard for interpretation.Three participants who did not meet the inclusion criteria were excluded based on the detection results,and a total of 639 participants were included in the study.The accuracy of the algorithm in identifying sinus tachycardia and rapid ventricular rate AF was evaluated using metrics such as recall rate,precision rate,macro F1 score for multi-class classification.Results Among 639 subjects,there were 469 males and 170 females.There were 389 cases of sinus tachycardia and 250 cases of rapid ventricular rate AF,with a mean age of(46.53±13.32)years.The recall rate,precision rate,and F1 value of iECG algorithm in identifying sinus tachycardia was 98.7%,99.2%and 99.0%,respectively,while it was 98.8%,98.0%and 98.4%,respectively for AF with rapid ventricular rate.The macro F1 of AF with rapid ventricular rate and sinus tachycardia was 98.7%.The iECG based on the intelligent wristwatch showed good consistency with the corresponding 12L-ECG waveforms.Conclusion The intelligent wristwatch-based iECG algorithm can effectively identify sinus tachycardia and rapid ventricular rate AF,demonstrating good ac

关 键 词:窦性心动过速 心房颤动 单导联心电图 快心室率 可穿戴设备 

分 类 号:R541.7[医药卫生—心血管疾病]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象