Reducing influence of doppler ultrasound signal processing techniques on fetal heart rate variability

Commonly used noninvasive fetal monitoring is aimed at evaluation of fetal well-being, which is done by variability analysis of fetal heart rate (FHR). The FHR signal is usually provided by fetal monitors which operate on the Doppler ultrasound (US) signal coming from the mechanical activity of the fetal heart. Although the measurement accuracy of today's fetal monitors is quite satisfactory comparing to the reference direct electrocardiography, the FHR variability indices obtained from US signal are considerably decreased, thus, their clinical value is rather low. The effect of underestimation of the FHR variability was recognized for a given type of fetal monitor as being related to an autocorrelation technique applied for periodicity measurement. In order to reduce this error we proposed a correction method which relies upon removing of the constant error component caused by averaging nature of the autocorrelation. The random error component is still present, but the reliability of short-term FHR variability assessment in global perspective (for example one-hour monitoring record) was significantly improved.