心电实时监护算法及监护系统研究
|
摘要
心脏病是一种常见的多发慢性疾病,严重危害人们健康。心脏病具有突发性,为了及时诊断,需要对心电进行监护,现在的实时监护大多局限于医院,这种监护既昂贵又给患者带来了不便。随着无线网络技术及微控制技术的发展,基于无线网络技术的心电实时监护系统也初见端倪,此类系统克服了医院实时监护的弊端,具有很高的实用价值和广阔的应用前景。
基于无线网络技术的心电系统由心电监护终端、无线通信网络和心电监护中心三大部分组成。心电监护终端具有心电实时检测功能,依靠对心电波形各个特征点的识别结果,根据医学知识进行诊断分类。目前,特征点检测的方法繁多,诸如:差分阈值法、模板匹配法、小波变换法及神经网络法等,这些算法在工程应用上存在着一些弊端。本文在分析上述算法的基础上,提出了一种新的心电实时检测算法:差分窗口法DWA,该算法依次对心电信号的几个主要特征参数进行提取和识别,从而能够快速准确地对心电图进行诊断,易于在便携式系统中实现,论文最后以DWA算法为核心,设计了心电实时监护终端。
论文的主要工作如下:
(1)在分析心电检测算法研究现状的基础上,提出了一种新的差分阈值法,即差分窗口法DWA,该算法首先用基线-相对幅值-差分法检测出QRS波群,再采用均值-方差-窗口法对TP波进行检测。仿真实验采用了MIT-BIH Arrhythmia Database中提供的48组心电数据及一些临床采集的心电数据,实验表明准确率高达99.98%。
(2)将DWA算法移植到ARM7平台上,设计了心电实时监护系统终端,并采用上述实验数据测试及临床试验,准确率高达99.96%,验证了系统的可行性和可靠性,实现了对心电的实时监测。
(3)对心电实时监护网络系统进行了分析,进行了节点、网络拓扑结构设计,并做了相关网络实验,为心电实时监护系统的网络设计作了前期实验工作。
The heart disease is a common multiple chronic disease which is serious for human health. The heart disease is hard to be diagnosed and ECG monitoring is vital for that. Real-time monitoring systems are now mostly servicing in hospitals, and that is too expensive and inconvenienced for the patients. As the developing of the wireless network technology and the micro-control technology, real-time ECG monitoring systems based on wireless network technology appears. These models of monitoring systems can overcome the shortcomings of the systems working in hospitals, and they are very practical and potential in future applications.
ECG system which based on wireless network technology consists of ECG terminal devices, wireless networks and electrocardiogram monitoring center. At present, there are lots of methods in feature-point detection area, such as: differential threshold, template matching, wavelet transform neural network algorithm. Of course defects in these engineering applications of the algorithms talked above are inevitable. After the analysis of the algorithms listed above, this dissertation presents a new real-time ECG detection algorithm: Derivative Window Algorithm(DWA). This algorithm can extract and recognize the parameters of the features in ECG signal in turn, and could make diagnosis of the electrocardiogram rapidly and accurately, and finally being able to implement in portable systems. At last, the dissertation designs a real-time ECG monitoring terminal based on DWA.
The specific content of this dissertation is as follows:
(1)Based on the current situation analysis of ECG detection algorithm, a novel differential threshold Algorithm named Derivative Window Algorithm(DWA) is presented. The algorithm first detect the QRS wave group with the baseline– opposite amplitude- derivative method, and then use the mean - variance - window method to detect the TP wave. The ECG data in the simulation experiments come from the MIT-BIH Arrhythmia Database and besides that some clinical too. The experiments show that the rate of accuracy is reach up to 99.98%.
(2) The terminal of real-time ECG monitoring system is designed through transplanting the DWA to ARM7 platform. And then using the experimental data and the clinical testing to identify the rate of accuracy(99.96%). The result verifies the feasibility and reliability of the system, and achieves the real-time monitoring of the electro cardio.
(3)analysing the real-time ECG monitoring system network, including the joint design and the network topology design. And at last, some related network experiments are made for the real-time ECG monitoring system network design.
基于无线网络技术的心电系统由心电监护终端、无线通信网络和心电监护中心三大部分组成。心电监护终端具有心电实时检测功能,依靠对心电波形各个特征点的识别结果,根据医学知识进行诊断分类。目前,特征点检测的方法繁多,诸如:差分阈值法、模板匹配法、小波变换法及神经网络法等,这些算法在工程应用上存在着一些弊端。本文在分析上述算法的基础上,提出了一种新的心电实时检测算法:差分窗口法DWA,该算法依次对心电信号的几个主要特征参数进行提取和识别,从而能够快速准确地对心电图进行诊断,易于在便携式系统中实现,论文最后以DWA算法为核心,设计了心电实时监护终端。
论文的主要工作如下:
(1)在分析心电检测算法研究现状的基础上,提出了一种新的差分阈值法,即差分窗口法DWA,该算法首先用基线-相对幅值-差分法检测出QRS波群,再采用均值-方差-窗口法对TP波进行检测。仿真实验采用了MIT-BIH Arrhythmia Database中提供的48组心电数据及一些临床采集的心电数据,实验表明准确率高达99.98%。
(2)将DWA算法移植到ARM7平台上,设计了心电实时监护系统终端,并采用上述实验数据测试及临床试验,准确率高达99.96%,验证了系统的可行性和可靠性,实现了对心电的实时监测。
(3)对心电实时监护网络系统进行了分析,进行了节点、网络拓扑结构设计,并做了相关网络实验,为心电实时监护系统的网络设计作了前期实验工作。
The heart disease is a common multiple chronic disease which is serious for human health. The heart disease is hard to be diagnosed and ECG monitoring is vital for that. Real-time monitoring systems are now mostly servicing in hospitals, and that is too expensive and inconvenienced for the patients. As the developing of the wireless network technology and the micro-control technology, real-time ECG monitoring systems based on wireless network technology appears. These models of monitoring systems can overcome the shortcomings of the systems working in hospitals, and they are very practical and potential in future applications.
ECG system which based on wireless network technology consists of ECG terminal devices, wireless networks and electrocardiogram monitoring center. At present, there are lots of methods in feature-point detection area, such as: differential threshold, template matching, wavelet transform neural network algorithm. Of course defects in these engineering applications of the algorithms talked above are inevitable. After the analysis of the algorithms listed above, this dissertation presents a new real-time ECG detection algorithm: Derivative Window Algorithm(DWA). This algorithm can extract and recognize the parameters of the features in ECG signal in turn, and could make diagnosis of the electrocardiogram rapidly and accurately, and finally being able to implement in portable systems. At last, the dissertation designs a real-time ECG monitoring terminal based on DWA.
The specific content of this dissertation is as follows:
(1)Based on the current situation analysis of ECG detection algorithm, a novel differential threshold Algorithm named Derivative Window Algorithm(DWA) is presented. The algorithm first detect the QRS wave group with the baseline– opposite amplitude- derivative method, and then use the mean - variance - window method to detect the TP wave. The ECG data in the simulation experiments come from the MIT-BIH Arrhythmia Database and besides that some clinical too. The experiments show that the rate of accuracy is reach up to 99.98%.
(2) The terminal of real-time ECG monitoring system is designed through transplanting the DWA to ARM7 platform. And then using the experimental data and the clinical testing to identify the rate of accuracy(99.96%). The result verifies the feasibility and reliability of the system, and achieves the real-time monitoring of the electro cardio.
(3)analysing the real-time ECG monitoring system network, including the joint design and the network topology design. And at last, some related network experiments are made for the real-time ECG monitoring system network design.
引文
[1]邦荃译.生物医学数字信号处理[M].武汉:华中科技大学出版社,2001:11,146页.
[2]王陈海,吴太虎.短距离无线通信技术发展及在医疗监护中的应用[J].医疗卫生装备,2008(01):30-34.
[3]苗建宁,朱宇. Holter技术发展展望[J].医学信息,2006,19(02):202-204.
[4]林天毅,段会龙,吕维雪.远程医疗信息系统的应用及相关问题.国外医学生物医学工程分册.1998,21(4):202-206.
[5]单红旭.兼具Holetr功能的数字式心电遥测监护系统的研制[J].仪器仪表学报.1999,20(4): 348-350.
[6] TORAL SL.QUERO JM.PEREZ EM, A microprocessor based system for ECG telemedi- cine and telecare[J]. Materials Research Society Symposium-Proceedings, 2001, 624(4):526-529.
[7]黄宛.临床心电图学[M].人民卫生出版社,1989:10-23.
[8]王志毅,高克俭.心电图形态诊断学[M].天津科学技术初版社,2001:20-35.
[9]吴宝明,朱凌云.远程心电监护系统及动态心电信号自动分析的研究进展[J].中国医疗器械新志,2006;30(5):352-355.
[10]王恒迪,朱坚民,梁丽,沈波.一种简单准确的R波检测算法.生物医学工程学杂志,2007;24(5):1019-1022.
[11]苏丽,赵国良,李东明.心电信号QRS波群检测算法研究[J].哈尔滨工程大学学报,2005;26(4): 513-517.
[12]王鸿鹏,戴博,杨孝宗.心电自动分析系统[J].北京生物医学工程,2007;26(4):386-389.
[13]朱凌云,吴宝明,王正国,李刚,曹长修,朱新建,闫庆广.移动心电监护系统QRS波的实时检测算法研究[J].仪器仪表学报,2005;26(6):603-607.
[14]涂承媛,曾衍钧,李树信.基于频数直方图检测QRS波的新方法[J].生物医学工程学杂志,2003,20(2):325-327.
[15]何欣,张永红,白净.基于小波变换模板的QRS检测方法[J].航天医学与医学工程,2003,(06).
[16]王磊,郑崇勋,叶继伦等.一种高效的QRS波实时检测方法[J].北京生物医学工程, 1998,17(4):217-222.
[17]邱雅竹,丁显峰等.基于Mexican-hat小波的QRS检测新方法[J].生物医学工程杂志, 2006;23(6):1347-1349
[18]段靖芳,郝莹.基于小波变换和神经网络的心电信号P波检测[J].生物医学工程研究, 2005;24(4):209-211.
[19]万相奎,秦树人,梁小容,丁建平.基于“小波-幅值-斜率”的P波检测新方法[J].生物医学工程杂志,2006;23(4):722-725.
[20]张永红等.基于“小波-相对幅度-曲率”的P波检测算法[J].中国医疗器械新志,2003;27(2):104-108.
[21]夏恒超,詹永麒,陈旭华.基于小波变换和波形信息的P波检测方法[J].北京生物医学工程,2003;22(1):27-29.
[22]季虎,孙即祥,王春光.基于小波变换的自适应QRS-T对消P[J].电子与信息学报,2007;29(8):1868-1871.
[23]王文,孙世双,周勇.基于小波变换的心电图QRS波群检测方法研究[J].北京生物医学工程,2002,21(4):241-247.
[24]丁哨卫,张作生,冯焕清.基于自适应小波变换的QRS波检测算法[J].中国科学技术大学学报,1998,28(5):580-586.
[25]王伟,李小梅.利用自适应小波变换检测心电QRS波算法的实现[J].科技通报, 2008,(01).
[26]陆英北,张增芳,蔡坤宝.基于小波变换的心电信号基线矫正方法[J].北京生物医学工程, 2000,(04).
[27]李章勇,刘圣蓉,谢正祥.利用小波变换提取和分析心动周期信号中子信号[J].重庆邮电学院学报(自然科学版) , 2006,(01)
[28]朱洪俊,秦树人,曹怡.心电信号中R波的小波探测法[J].重庆大学学报(自然科学版),2002,(05).
[29]季虎,毛玲,孙即祥.基于小波变换与形态学运算的R波检测算法[J].计算机应用,2006,26(5):1223-1225.
[30]赵志华.基于形态小波的QRS波检测算法[J].计算机工程与应用.2007,43(13).
[31]刘少颖,卢继来,郝丽,胡广书基于数学形态学和小波分解的QRS波群检测算法,清华大学学报(自然科学版),2004,44(6)852:854.
[32]郑玲,叶大田,杨福生.数学形态学及其在生物医学中的应用[J].国外医学生物医学工程分册,1994,17(3):125-136.
[33]高艳,胡阳.基于数学形态学方法的心电图波形分离技术[J].生物医学工程学杂志,2001,18(1):55-59.
[34]周开利,康耀红.神经网络模型及其MATLAB仿真程序设计[M].北京:清华大学出版社,2005.
[35]于学鸿,徐小汉.基于神经网络的波形检测方法[J].生物医学工程学杂志,2000,17(1):59-62.
[36]杨军,王宏山,俞梦孙.心电图ST段测量的神经网络方法[J].北京生物医学工程,2002,(02).
[37]冯俊,邱雅竹,莫智文.基于小波特征提取的多导联心电图神经网络分类[J].第三军医大学学报,2006,(08).
[38]Szilagyi SM,Szilagyi L,David L.Comparison between neural-network-based adaptive filter- ing and wavelet transform for ECG characteristic points detection[C].Proceedings-19th Inter-national Conference-IEEE/EMBS,Chicago,IL.USA.1997.272-274.
[39] LPC2214/2124/2212/2214 User Manual[R], Preliminary Release. PHILIPS, 2004.
[40]周立功.深入浅出ARM7:LPC213X/214X(上册)[M].北京:北京航天航空大学出版社,2005.
[41]金敏,周翔,金梁.嵌入式系统-组成原理与设计编程[M].北京:人民邮电出版社,2006.
[42]金纯,罗祖秋,罗凤,陈前斌.ZigBee技术基础及案例分析[M].北京:国防工业出版社.2008.
[43]瞿雷,刘盛德,胡咸斌.ZigBee技术及应用[M].北京:北京航空航天大学出版社.2007
[44] Wheeler Ander. Commercial applications of wireless sensor networks using ZigBee[J]. IEEE CommunicationsMagazine, 2007,45(4):70-77.
[45] StanislavSafarie,Kresimir Malarie ZigBee wireless standard 48th Intemational SymPosium ELMAR-2006:259-262.
[46] ZigBee Specification Version1.1,september:5th 2005.
[47] IEEE802.15.4,Part 15.4: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low-Rate Wireless Personal Area Network (LR-WPANs).October,2003.
[48]徐勇军,刘峰,王春芳,姜鹏.低速无线个域网实验教程[M].北京:北京理工大学出版社.2008.
[49]Chipcon.2004.CC2420 data sheet.
[50]代忠,樊晓光,万明,周灼荣.基于射频芯片CC2420的ZigBee无线通信节点设计[J].传感器世界,2006,5:41-44.
[51] Preliminary-JN-DS-JN5139-xxx-Myy v1.2, Jennic,2007.
[2]王陈海,吴太虎.短距离无线通信技术发展及在医疗监护中的应用[J].医疗卫生装备,2008(01):30-34.
[3]苗建宁,朱宇. Holter技术发展展望[J].医学信息,2006,19(02):202-204.
[4]林天毅,段会龙,吕维雪.远程医疗信息系统的应用及相关问题.国外医学生物医学工程分册.1998,21(4):202-206.
[5]单红旭.兼具Holetr功能的数字式心电遥测监护系统的研制[J].仪器仪表学报.1999,20(4): 348-350.
[6] TORAL SL.QUERO JM.PEREZ EM, A microprocessor based system for ECG telemedi- cine and telecare[J]. Materials Research Society Symposium-Proceedings, 2001, 624(4):526-529.
[7]黄宛.临床心电图学[M].人民卫生出版社,1989:10-23.
[8]王志毅,高克俭.心电图形态诊断学[M].天津科学技术初版社,2001:20-35.
[9]吴宝明,朱凌云.远程心电监护系统及动态心电信号自动分析的研究进展[J].中国医疗器械新志,2006;30(5):352-355.
[10]王恒迪,朱坚民,梁丽,沈波.一种简单准确的R波检测算法.生物医学工程学杂志,2007;24(5):1019-1022.
[11]苏丽,赵国良,李东明.心电信号QRS波群检测算法研究[J].哈尔滨工程大学学报,2005;26(4): 513-517.
[12]王鸿鹏,戴博,杨孝宗.心电自动分析系统[J].北京生物医学工程,2007;26(4):386-389.
[13]朱凌云,吴宝明,王正国,李刚,曹长修,朱新建,闫庆广.移动心电监护系统QRS波的实时检测算法研究[J].仪器仪表学报,2005;26(6):603-607.
[14]涂承媛,曾衍钧,李树信.基于频数直方图检测QRS波的新方法[J].生物医学工程学杂志,2003,20(2):325-327.
[15]何欣,张永红,白净.基于小波变换模板的QRS检测方法[J].航天医学与医学工程,2003,(06).
[16]王磊,郑崇勋,叶继伦等.一种高效的QRS波实时检测方法[J].北京生物医学工程, 1998,17(4):217-222.
[17]邱雅竹,丁显峰等.基于Mexican-hat小波的QRS检测新方法[J].生物医学工程杂志, 2006;23(6):1347-1349
[18]段靖芳,郝莹.基于小波变换和神经网络的心电信号P波检测[J].生物医学工程研究, 2005;24(4):209-211.
[19]万相奎,秦树人,梁小容,丁建平.基于“小波-幅值-斜率”的P波检测新方法[J].生物医学工程杂志,2006;23(4):722-725.
[20]张永红等.基于“小波-相对幅度-曲率”的P波检测算法[J].中国医疗器械新志,2003;27(2):104-108.
[21]夏恒超,詹永麒,陈旭华.基于小波变换和波形信息的P波检测方法[J].北京生物医学工程,2003;22(1):27-29.
[22]季虎,孙即祥,王春光.基于小波变换的自适应QRS-T对消P[J].电子与信息学报,2007;29(8):1868-1871.
[23]王文,孙世双,周勇.基于小波变换的心电图QRS波群检测方法研究[J].北京生物医学工程,2002,21(4):241-247.
[24]丁哨卫,张作生,冯焕清.基于自适应小波变换的QRS波检测算法[J].中国科学技术大学学报,1998,28(5):580-586.
[25]王伟,李小梅.利用自适应小波变换检测心电QRS波算法的实现[J].科技通报, 2008,(01).
[26]陆英北,张增芳,蔡坤宝.基于小波变换的心电信号基线矫正方法[J].北京生物医学工程, 2000,(04).
[27]李章勇,刘圣蓉,谢正祥.利用小波变换提取和分析心动周期信号中子信号[J].重庆邮电学院学报(自然科学版) , 2006,(01)
[28]朱洪俊,秦树人,曹怡.心电信号中R波的小波探测法[J].重庆大学学报(自然科学版),2002,(05).
[29]季虎,毛玲,孙即祥.基于小波变换与形态学运算的R波检测算法[J].计算机应用,2006,26(5):1223-1225.
[30]赵志华.基于形态小波的QRS波检测算法[J].计算机工程与应用.2007,43(13).
[31]刘少颖,卢继来,郝丽,胡广书基于数学形态学和小波分解的QRS波群检测算法,清华大学学报(自然科学版),2004,44(6)852:854.
[32]郑玲,叶大田,杨福生.数学形态学及其在生物医学中的应用[J].国外医学生物医学工程分册,1994,17(3):125-136.
[33]高艳,胡阳.基于数学形态学方法的心电图波形分离技术[J].生物医学工程学杂志,2001,18(1):55-59.
[34]周开利,康耀红.神经网络模型及其MATLAB仿真程序设计[M].北京:清华大学出版社,2005.
[35]于学鸿,徐小汉.基于神经网络的波形检测方法[J].生物医学工程学杂志,2000,17(1):59-62.
[36]杨军,王宏山,俞梦孙.心电图ST段测量的神经网络方法[J].北京生物医学工程,2002,(02).
[37]冯俊,邱雅竹,莫智文.基于小波特征提取的多导联心电图神经网络分类[J].第三军医大学学报,2006,(08).
[38]Szilagyi SM,Szilagyi L,David L.Comparison between neural-network-based adaptive filter- ing and wavelet transform for ECG characteristic points detection[C].Proceedings-19th Inter-national Conference-IEEE/EMBS,Chicago,IL.USA.1997.272-274.
[39] LPC2214/2124/2212/2214 User Manual[R], Preliminary Release. PHILIPS, 2004.
[40]周立功.深入浅出ARM7:LPC213X/214X(上册)[M].北京:北京航天航空大学出版社,2005.
[41]金敏,周翔,金梁.嵌入式系统-组成原理与设计编程[M].北京:人民邮电出版社,2006.
[42]金纯,罗祖秋,罗凤,陈前斌.ZigBee技术基础及案例分析[M].北京:国防工业出版社.2008.
[43]瞿雷,刘盛德,胡咸斌.ZigBee技术及应用[M].北京:北京航空航天大学出版社.2007
[44] Wheeler Ander. Commercial applications of wireless sensor networks using ZigBee[J]. IEEE CommunicationsMagazine, 2007,45(4):70-77.
[45] StanislavSafarie,Kresimir Malarie ZigBee wireless standard 48th Intemational SymPosium ELMAR-2006:259-262.
[46] ZigBee Specification Version1.1,september:5th 2005.
[47] IEEE802.15.4,Part 15.4: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low-Rate Wireless Personal Area Network (LR-WPANs).October,2003.
[48]徐勇军,刘峰,王春芳,姜鹏.低速无线个域网实验教程[M].北京:北京理工大学出版社.2008.
[49]Chipcon.2004.CC2420 data sheet.
[50]代忠,樊晓光,万明,周灼荣.基于射频芯片CC2420的ZigBee无线通信节点设计[J].传感器世界,2006,5:41-44.
[51] Preliminary-JN-DS-JN5139-xxx-Myy v1.2, Jennic,2007.