【摘要】 目的:针对胎儿瞬时心率难以准确测量的问题,提出一种基于时频域的线性胎儿心音信号瞬时心率检测算法。方法:将临床采集到的胎儿心音信号进行预处理并进行短时傅里叶变换获得其时频图;选取胎儿心音信号模板和噪声信号模板;确定线性判别方程和判别常数;利用线性判别对胎儿心音信号进行定位,并计算胎儿瞬时心率。结果:经对医院采集的多例胎儿心音信号的测试表明,该算法的识别准确率高,计算速度快,且具有普遍适用性和有效性。结论:该算法解决了胎儿心音幅度小、易受噪声干扰的缺点,能够实时、快速、准确计算胎儿瞬时心率,并且可实现胎儿心音的定位,可为胎儿心音分析和胎儿监护提供良好的基础。更多还原

【Abstract】 Objective To develop a fetal cardiac sound instantaneous heart rate detection algorithm based on time-frequency domain and linear analysis. Methods Fetal heart sound sound signals underwent pretreat and short-time Fourier transform to obtain the time-frequency planes. Heart sound feature template and noise sound feature template were selected, and then linear discriminant equation and dicriminant constant were determined to locate the heart sound signal and calculate the instantaneous heart rate. Results The trials in multiple fetal heart sound signals acquired in the hospital proved that the algorithm behaved well in accuracy, velocity, adaptability and effectiveness. Conclusion The algorithm proposed can be used to calculate fetal instantaneous heart rate and locate the heart sound although the fetal heart sound is low-amplification and vulnerable to noise.更多还原