Estimated Fetal Weight Calculator

No ultrasound measures fetal mass directly. Estimated fetal weight is a regression prediction: measured body dimensions are entered into an equation fitted to the logarithm of birth weight in a reference population. The Hadlock four-parameter model uses biparietal diameter, head circumference, abdominal circumference, and femur length in centimeters, and its cross-product terms - 0.00061 times BPD times AC, and minus 0.00386 times AC times FL - exist because the relationship between linear dimensions and volume is not additive. Abdominal circumference carries the most predictive weight because it captures liver size and subcutaneous fat, the tissues that vary most with nutritional status, and it is measured at the level of the stomach and the junction of the umbilical vein with the portal sinus. Roughly 95 percent of estimates fall within plus or minus 15 percent of actual birth weight, with systematic error usually under 5 percent, but precision degrades below 1,000 g and above 4,000 g and when maternal BMI exceeds about 35. A fetus below the 10th percentile is small for gestational age and one above the 90th is large for gestational age; macrosomia is defined at 4,000 or 4,500 g depending on the source. ACOG does not support cesarean delivery for suspected macrosomia alone unless the estimate exceeds 5,000 g without diabetes or 4,500 g with diabetes, since positive predictive value for macrosomia is only about 40 to 60 percent. Serial scans spaced at least two weeks apart give growth velocity, which is more informative than any single estimate.

Attempts to weigh a fetus by ultrasound began soon after B-mode imaging entered obstetrics. James Willocks in Glasgow used biparietal diameter in the mid-1960s to estimate fetal size, but head measurement alone tracks skeletal growth rather than soft tissue and performed poorly at the extremes. Stuart Campbell and Diana Wilkin published a decisive advance in 1975 in the British Journal of Obstetrics and Gynaecology, showing that abdominal circumference alone predicted birth weight substantially better than head measurements, because the fetal abdomen reflects hepatic glycogen and fat stores. Two-parameter equations followed, among them the widely cited formula of Michael Shepard and colleagues in 1982 combining biparietal diameter and abdominal circumference. Frank Hadlock and coworkers at the University of Texas Health Science Center at Houston then published a family of regression models in Radiology in 1984 and 1985, derived from scans of well-dated pregnancies, showing that combining head, body, and femur measurements reduced random error relative to any two-parameter model. The four-parameter version became the default built into ultrasound machines worldwide. More recently, prescriptive international standards from the INTERGROWTH-21st Project and the World Health Organization fetal growth study, published during the 2010s, have addressed a separate question: which reference population a given fetus should be compared against.