Ultrasound Pregnancy Images Classification Using Artificial Neural Network

Abstract

This research introduces an automatic ultrasound human pregnancy images classification (male or female) system using artificial neural network to classify the pregnancy images. An automatic ultrasound human pregnancy images system consists of three Modules: The first is preprocessing ultrasound images (noise removing , image normalization and segmentation), second is feature extraction module then pregnancy classification module. After preprocessing, features extracting by using kernel principal component analysis (kernel PCA) after that, Elman neural network is used to a classify training and testing these pregnancy images. The system produces promising results for pregnancy images classification.