Effect of measurement errors on sonographic evaluation of ventriculomegaly

Abstract

Ventriculomegaly is a sensitive indicator of maldevelopment of the fetal brain and spinal cord. The fetal ventricular atrium is an optimal portion of the lateral ventricular system to measure in order to judge ventriculomegaly. We tested the susceptibility of this measurement to inaccuracies created by errors in cursor placement or improper selection of the plane of section for measurement in 52 normal fetuses. This measurement was shown to be reasonably insensitive to errors in cursor placement along the axis of the ventricular system. However, choice of an off-axis image plane of section, an angled measurement, or improper choice of ventricular boundary leads to a relatively large number of false-positive test results (approximately 10%). None of the simulated errors in atrial measurements underestimates the size of the ventricle, and thus, they are unlikely to result in a diminution of test sensitivity.