Dawes G S, Visser G H, Goodman J D, Redman C W
Am J Obstet Gynecol. 1981 Sep 1;141(1):43-52. doi: 10.1016/0002-9378(81)90673-6.
A method is described for the computerized numerical analysis of fetal heart periods (pulse intervals). It uses a digital filter to separate the record into its high- and low-frequency components and, after removal of baseline variation, identifies accelerations and decelerations of all sizes. It provides an objective method for separating episodes of high heart period variation, normally associated with fetal movements, from episodes of low variation. When Doppler ultrasound is used in the last 10 weeks of gestation, failure time averages 40%. Signal loss is not randomly distributed; it is on average 75% greater during episodes of high heart period variation, although it is not particularly associated with fetal movements as identified by nurse or patient. Nevertheless a comparison of simultaneous direct ECG and ultrasound records shows that the latter provide reasonable statistical measures of heart period variation, and also of accelerations and decelerations provided that signal loss is taken into account. The system thus provides a particularly useful adjunct to the analysis of antenatal human fetal heart rate records.
