论文部分内容阅读
胎动监护是预防围产期胎死的一项重要措施 .针对胎动发生时引起母体腹壁压力改变这一生理特性 ,首先利用压力传感器从母腹部提取胎动信号并经采样至计算机内 ;然后针对腹部胎动信号成分复杂、信噪比较低的特点 ,采用数字低通滤波和微分、积分技术对其进行预处理 ,经分析后提取其特征参量 ;最后利用贝叶斯最小错误率决策准则进行模式识别 .通过对 13例共 2 6 0分钟的腹部胎动信号的分析、处理及分类 ,其正确识别率达 84.35 % .
Fetal movement monitoring is an important measure to prevent perinatal fetal death.Aiming at the physiological characteristics that cause the change of maternal abdominal wall pressure when fetal movement occurs, the fetus signal is fetched from the mother’s abdomen using a pressure sensor and sampled into the computer. Signal complexity, low signal-to-noise ratio, preprocessing by digital low-pass filtering, differential and integral techniques, extracting the characteristic parameters after analysis, and finally using the Bayesian minimum error rate decision criterion for pattern recognition. Through 13 cases of a total of 260 minutes of fetal fetal signal analysis, processing and classification, the correct recognition rate of 84.35%.