一种改进的子宫肌电包络提取算法

发布时间：2018-08-30 14:57

【摘要】：基于腹部采集的子宫肌电信号(EMG)提取子宫收缩信息被认为是一种最有可能取代分娩力描记法(TOCO)的新方法。传统均方根(RMS)算法对子宫肌电信号中脉冲性噪声的抑制能力有限,为了有效克服上述问题,本文提出一种改进的子宫肌电包络提取算法。新方法先对原始子宫肌电信号进行过零检测,识别出爆发波并分段,再对分段后的信号进行不同窗口宽度的平滑滤波,最后采用均方根法提取子宫肌电强度(IEMG)作为包络。为了评估本算法的性能表现,本文将该算法与两种现有子宫肌电强度提取算法进行比较。结果表明,该算法在抑制子宫肌电信号中存在的脉冲性噪声方面要明显优于对比算法,其测量灵敏度与阳性预测值(PPV)分别为0.952和0.922,不仅明显高于对比算法一中对应参数的0.859和0.847,同时也高于对比算法二中的0.928和0.877。本研究表明了该方法的可靠性与有效性。
[Abstract]:The extraction of uterine contraction information by (EMG) based on abdominal electromyography (EMG) is considered to be a new method which is most likely to replace (TOCO). The traditional root mean square (RMS) (RMS) algorithm has a limited ability to suppress impulsive noise in electromyography (EMG) signals. In order to overcome the above problems, an improved EMG envelope extraction algorithm is proposed in this paper. The new method firstly detects the original electromyoelectric signals across zero, then recognizes the burst waves and segments, then smoothes the segmented signals with different window widths. Finally, the (IEMG) of uterine myoelectric intensity is extracted by the root mean square method as the envelope. In order to evaluate the performance of this algorithm, the algorithm is compared with two existing methods of myoelectric strength extraction. The results show that the proposed algorithm is superior to the contrast algorithm in suppressing the impulse noise in the electromyography signals. The measured sensitivity and positive predictive value (PPV) were 0.952 and 0.922, respectively, which were significantly higher than those in comparison algorithm 1 (0.859 and 0.847) and in contrast algorithm 2 (0.928 and 0.877). This study shows the reliability and effectiveness of the method.
【作者单位】：暨南大学电子工程系;
【基金】：国家国际科技合作专项(No.2015DFI12970) 广东省科技计划应用型科技研发专项资金项目(No.2015B020233010) 广州市科技计划产学研协同创新重大专项(No.201508030008)
【分类号】：R714;TN911.7

【参考文献】