COMPARISON OF MORPHOLOGICAL-BASED SEGMENTATION METHODS FOR FETAL FEMUR LENGTH MEASUREMENTS

Abstract

Measurement of fetal femur length is crucial for the estimation of fetal age and growth pattern. Due to the noisy nature of ultrasound images and the variation in image acquisition and measurement techniques, manual measurements are subject to interobserver and intraobserver variability. The motivation of this paper is to study the efficiency of automated methods used in the femur endpoint detection module of our proposed computer-assisted processing and analysis system, which is employed to determine the gestational age for promoting the standardization of fetal femur measurements, thus resulting in more reliable age estimates. The proposed methods are based on the morphological operators methods5and the morphological watershed segmentation method. We applied these methods on a set of 100 digitized fetal femur images of size 287 × 224 pixels with 256 gray-level images and a resolution of 100 pixels per inch (ppi) horizontally and vertically. The results showed that the watershed method was the more efficient one and gave reliable results for every femur image tested.