Estimation of Fetal Weight by Single Multiplicative Neuron Model Using Symphysis-Fundus Height

摘要

Estimation of fetal weight is used for maternity care management, especially in counseling, differential diagnoses, detection of intrauterine growth retardation and the mode of delivery. The aim of the present study was to evaluate the accuracy of using single multiplicative neuron model for estimation of fetal weight after 24 weeks gestation. Four-hundred and sixty-seven patients with a healthy, singleton pregnancy, whom fetal biometry measurements were routinely performed by ultrasound (US), considered as the training group. 4 input variables were used to construct single multiplicative neuron model: abdominal circumference (AC), head circumference (HC), femur length (FL) and symphysis-fundus height (SFH). AC, HC, FL parameters were measured by US and SFH parameter measure clinically. Also a total of 205 fetuses were assessed subsequently as the validation group. In the validation group the mean absolute error and the mean absolute percent error was 190.22 g and 6.89% respectively. In conclusion, using symphisis-fundus height within single multiplicative neuron model could provide acceptable US estimation of fetal weight.