面向胎心电R波检测的若干研究

发布时间：2018-03-12 08:29

本文选题：胎心电　切入点：R波检测　出处：《南京大学》2014年硕士论文　论文类型：学位论文

【摘要】：目前,胎儿心电R波检测技术还不成熟,尚不能满足胎儿心电波形提取或胎儿瞬时心率计算的客观要求。之所以如此,一个重要原因是,我们检测的孕妇腹壁电信号,是一个比常规成人ECG信号复杂得多的混合信号,包含了母体心电成分、胎儿心电成分、还有其他干扰和噪声等。概括起来,胎儿心电R波检测存在的问题主要体现在两方面：一、检测的腹壁电信号的胎心电信噪比太低,在去除了母体心电成分的残留信号中胎儿心电成分被埋没在强背景噪声中,难于正确检测出胎儿心电R波波峰位置；二、检测的腹壁电信号中母体心电成分过弱或者胎儿心电成分过强,则母体心电成分的R波波峰不能直接正确检测,最终又影响到胎儿心电成分的R波检测。本论文就是面向上述两方面的胎儿心电R波检测问题开展研究,具体研究工作如下：(1)进行实际孕妇腹壁电数据的检测,构建自己的孕妇腹壁电数据库。(2)基于实测数据,研究了电极如何布置才能使获得的腹壁电信号有较高的胎心电信号信噪比的问题,找到了几组电极布置的优势方向,得出了优势方向电极布置间距优化的定性结论,并给出了解释。(3)对个别检测通道信号显示出较高母体ECG信号信噪比的情况,容易正确检测其母体R波波峰,可用此R波波峰检测结果作为参考,帮助其他通道进行母体R波波峰的正确检测,即在参考位置的附近进行局部搜索即可。由于局部自动搜索需要一个范围,为此,本文基于课题组实测数据和MIT Non-Invasive Fetal Electrocardiogram Database数据,研究了各通道信号对应R波波峰位置的差异规律,得出了差异范围不超过30 ms的结论。(4)针对各通道胎心电普遍明显进而影响腹壁电母体R波波峰正确检测的问题,一方面,优化了单通道信号小波分解重构方法,提高了单通道母体R波波峰正确检测率；另一方面,提出了一个基于各通道R波检测结果对多检漏检进行融合判断的方法,进一步提高了母体R波波峰的检测正确率。全文共分6章,第1章为绪论,第2-5章分别对应以上4项工作,第6章给出了本文工作的总结,并对下一步的工作给予了展望。
[Abstract]:At present, fetal ECG R wave detection technology is not mature and can not meet the objective requirements of fetal ECG waveform extraction or fetal instantaneous heart rate calculation. Is a mixture of signals that are much more complex than conventional adult ECG signals, including maternal ECG, fetal ECG, and other disturbances and noises. The problems of fetal ECG R wave detection are mainly reflected in two aspects: first, the fetal ECG signal to noise ratio of the detected abdominal wall electrical signal is too low, in removing the residual signals of maternal ECG components, fetal ECG components are buried in strong background noise. It is difficult to correctly detect the position of the peak of fetal ECG R wave. Secondly, if the maternal ECG component is too weak or the fetal ECG component is too strong, the R wave peak of the maternal ECG component can not be detected directly and correctly. This paper is aimed at the above two aspects of fetal ECG R wave detection research, the specific research work is as follows: 1) the actual pregnant women's abdominal wall electrical data detection. Based on the measured data, this paper studies how to arrange the electrodes in order to make the obtained abdominal wall electrical signals have higher signal-to-noise ratio (SNR) of fetal ECG signals, and find out the superior directions of several groups of electrodes. The qualitative conclusion of optimizing the spacing of dominant direction electrode arrangement is obtained, and the explanation is given. The signal to noise ratio (SNR) of individual detection channel signal is higher than that of parent ECG signal, and it is easy to detect the R wave peak of parent body correctly. We can use this R wave peak detection result as a reference to help other channels to detect the parent R wave peak correctly, that is, to make a local search near the reference position. Based on the measured data and MIT Non-Invasive Fetal Electrocardiogram Database data, the differences of the position of the R wave peak corresponding to each channel signal are studied in this paper. The conclusion that the range of difference is not more than 30 Ms is obtained. Aiming at the problem that every channel fetal ECG is obviously obvious and then affecting the correct detection of the peak of R wave of the abdominal wall electric mother body, on the one hand, the method of wavelet decomposition and reconstruction of single channel signal is optimized. On the other hand, a fusion judgment method based on the R wave detection results of each channel is proposed. The paper is divided into 6 chapters, the first chapter is the introduction, the 2-5 chapters correspond to the above four tasks, the sixth chapter gives the summary of the work, and gives the prospects for the next work.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R714.5;TN911.7

【相似文献】