基于改进差分阈值算法的心电检测技术研究

doi:10.16180/j.cnki.issn1007-7820.2020.02.002

摘要/Abstract

摘要：

针对差分阈值算法中固定阈值的局限性,文中提出了一种基于自适应波峰阈值和R波间隔阈值的算法。该算法结合心电信号特点自动选择波峰阈值,并选择R波间隔阈值,提高了算法的自适应性和准确率。文中以MIT-BIH心律失常数据库中的心电信号作为实验样本,采用带通滤波与小波阈值滤波相结合的方法完成心电信号去噪,采用改进差分自适应阈值算法对心电信号进行波形检测。实验结果表明,该算法能够将心电信号R波的检测准确率提升到99.57%,有效减少了误检、漏检问题的发生,并可准确完成心率、心率变异性、身体疲劳度、精神疲劳度计算和常见心律失常分类。

关键词: 小波阈值滤波, 改进差分阈值, 自适应, 心电检测, 人体生理状态, 心律失常分类

Abstract:

Aiming at the limitation of fixed threshold in differential threshold algorithm, an algorithm based on adaptive peak threshold and R wave interval threshold was proposed. The algorithm automatically selected the peak threshold based on the characteristics of the ECG signal, and selected the R-wave interval threshold to improve the adaptability and accuracy of the algorithm. In this study, the ECG signal in the MIT-BIH arrhythmia database was used as the experimental sample. The combination of bandpass filtering and wavelet threshold filtering were utilized to complete the denoising of ECG signals. The ECG signals were detected by the improved differential adaptive threshold algorithm. The experimental results showed that the algorithm could improve the detection accuracy of R wave of ECG signal to 99.57%. The algorithm effectively reduced the occurrence of false detections and missed inspections, and accurately calculated heart rate, heart rate variability, physical fatigue, mental fatigue and common arrhythmia classification.

Key words: wavelet threshold filtering, improved differential threshold, adaptive, ECG detection, human physiological state, classification of arrhythmias

中图分类号:

TN911.6

张晓军,吴芝路. 基于改进差分阈值算法的心电检测技术研究[J]. 电子科技, 2020, 33(2): 6-13.

ZHANG Xiaojun,WU Zhilu. Research of ECG Detection Technology Based on Improved Differential Threshold Algorithm[J]. Electronic Science and Technology, 2020, 33(2): 6-13.

图/表 16

表1

图1

表2

图2

表3

图3

图4

图5

图6

图7

表4

表5

图8

图9

图10

图11

参考文献 16

[1]	李中健, 李世锋, 申继红 , 等. 心电图学系列讲座(三)-心电图一般知识[J]. 中国全科医学, 2014,17(3):360-362.
	Li Zhongjian, Li Shifeng, Shen Jihong , et al. ECG lecture series (3)-general knowledge of ECG[J]. Chinese General Medicine, 2014,17(3):360-362.
[2]	Sörnmo L, Laguna P. Bioelectrical signal processing in cardiac and neurological applications[M]. Cambridge: Elsevier Academic Press, 2005.
[3]	Karmakar C, Jelinek H F, Khandoker A , et al. Multi-lag HRV analysis discriminates disease progression of post-infarct people with no diabetes versus diabetes[C]. Milan:Engineering in Medicine and Biology Society, 2015.
[4]	李延军, 严洪, 杨向林 , 等. 基于心率变异性的精神疲劳的研究[J]. 中国生物医学工程学报, 2010,29(1):1-6.
	Li Yanjun, Yan Hong, Yang Xianglin , et al. Mental fatigue based on heart rate variability[J]. Chinese Journal of Biomedical Engineering, 2010,29(1):1-6.
[5]	付川云 . 疲劳状态下驾驶人生理及眼动特征研究[D]. 哈尔滨:哈尔滨工业大学, 2011.
	Fu Chuanyun . Study on physiological and eye movement characteristics of drivers under fatigue[D]. Harbin:Harbin Institute of Technology, 2011.
[6]	李宏恩 . 心电信号检测中滤除肌电干扰的方法[J]. 电子科技, 2014,27(2):66-67,70.
	Li Hongen . Method of filtering ECG interference in ECG signal detection[J]. Electronic Science and Technology, 2014,27(2):66-67,70.
[7]	尚宇, 徐婷, 何永辉 . 分数阶傅里叶变换在心电信号处理中的应用[J]. 电子科技, 2011,24(8):116-118.
	Shang Yu, Xu Ting, He Yonghui . Application of fractional Fourier transform in ECG signal processing[J]. Electronic Science and Technology, 2011,24(8):116-118.
[8]	余顺 . 心电图波形的形态学识别[D]. 金华:浙江师范大学, 2014.
	Yu Shun . Morphological recognition of ECG waveform[D]. Jinhua:Zhejiang Normal University, 2014.
[9]	Sahoo S K, Subudhi A K, Kanungo B , et al. Notice of retraction feature extraction of ECG signal based on wavelet transform for arrhythmia detection[C]. Visakhapatnam:International Conference on Electrical, Electronics, Signals,Communication and Optimization, 2015.
[10]	Chauhan S, Vig L . Anomaly detection in ECG time signals via deep long short-term memory networks[C]. Paris:IEEE International Conference on Data Science and Advanced Analytics, 2015.
[11]	黄家洺, 郑日荣, 曾绳涛 . 心电信号特征检测算法研究[J]. 机电工程技术, 2013(8):106-110.
	Huang Jiaming, Zheng Rirong, Zeng Shutao . Research on the detection algorithm of ECG signal characteristics[J]. Mechatronics Engineering Technology, 2013(8):106-110.
[12]	陈耿铎, 曾有灵, 李喆 . 自适应双阈值心电信号检测算法研究[J]. 暨南大学学报(自然科学与医学版), 2018,191(3):81-87.
	Chen Gengduo, Zeng Youling, Li Zhe . Research on adaptive dual-threshold ECG signal detection algorithm[J]. Journal of Jinan University(Natural Science and Medicine Edition), 2018,191(3):81-87.
[13]	赵羿欧, 刘扬 . 一种改进的差分阈值心电检测算法[J]. 计算机工程, 2011(S1):347-348.
	Zhao Yiou, Liu Yang . An improved differential threshold ECG detection algorithm[J]. Computer Engineering, 2011(S1):347-348.
[14]	王娟, 严生梅, 彭晓珊 . 基于MATLAB的MIT-BIH心电信号快速识别[J]. 科技创新与应用, 2018(8):32-33,35.
	Wang Juan, Yan Shengmei, Peng Xiaoshan . Mit-bih ECG signal fast recognition based on MATLAB[J]. Science and Technology Innovation and Application, 2018(8):32-33,35.
[15]	宋喜国, 邓亲恺 . MIT-BIH心率失常数据库的识读及应用[J]. 中国医学物理学杂志, 2004,21(4):230-232.
	Song Xiguo, Deng Qinkai . Recognition and application of MIT-BIH arrhythmia database[J]. Chinese journal of medical physics, 2004,21(4):230-232.
[16]	Tan H G R, Tan A C, Khong P Y , et al. Best wavelet function identification system for ECG signal denoise applications[C]. Kuala Lumpur:International Conference on Intelligent and Advanced Systems, 2007.

分解尺度	低频频域/Hz	高频频域/Hz
1	0~90	90~180
2	0~45	45-90
3	0~22.5	22.5~45
4	0~11.25	11.25~22.5
5	0~5.625	5.625~11.25
6	0~2.812 5	2.812 5~5.625
7	0~1.406 25	1.406 25~2.812 5
8	0~0.703 125	0.703 125~1.406 25
9	0~0.351 562 5	0.351 562 5~0.703 125

类别	判定条件
漏搏	2.4 s≥RR间期≥2×AHR
停搏	RR间期≥2.4 s
心动过速	IHR≥100次/min
心动过缓	IHR≤60次/min
心律不齐	相邻RR间期差≥0.16 s

记录编号	实际心拍数	检测心拍数	漏检数	准确率 /%
100	2 273	2 272	1	99.96
101	1 865	1 861	4	99.79
102	2 187	2 094	93	95.75
103	2 084	2 072	12	99.42
104	2 229	566	1 663	25.39
105	2 572	1 385	1 187	53.85
106	2 027	1 090	937	53.77
107	2 137	2 047	90	95.79
109	2 532	2 299	233	90.80
111	2 124	451	1 673	21.23
112	2 539	2 527	12	99.53
113	1 795	1 795	0	100
114	1 879	840	1 039	44.70
115	1 953	1 953	0	100
116	2 412	1 904	508	78.94
117	1 535	1 534	1	99.94
118	2 288	2 206	82	96.42
119	1 987	1 982	5	99.75
小计	38 418	30 878	7 540	80.84

记录编号	实际心拍数	检测心拍数	漏检数	误检数	准确率 /%
100	2 273	2 273	0	0	100
101	1 865	1 875	0	10	99.46
102	2 187	2 187	0	0	100
103	2 084	2 084	0	0	100
104	2 229	2 256	0	27	98.79
105	2 572	2 569	3	0	99.88
106	2 027	2 027	0	0	100
107	2 137	2 137	0	0	100
109	2 532	2 518	14	0	99.45
111	2 124	2 136	0	12	99.44
112	2 539	2 508	31	0	98.78
113	1 795	1 795	0	0	100
114	1 879	1 903	0	24	98.72
115	1 953	1 953	0	0	100
116	2 412	2 396	16	0	99.34
117	1 535	1 552	0	17	98.89
118	2 288	2 279	9	0	99.61
119	1 987	1 989	0	2	99.90
小计	38 418	38 437	73	92	99.57