Detection and Diagnosis of Fetus Ultra Sound Images using Deep Learning Classification Methods

This research work proposes detection and diagnosis of Down syndrome using soft computing and deep learning algorithm. The detection module of the proposed DS classification system consists of Gabor transform, feature extractions and Convolutional Neural Networks (CNN) classification methods. The suggested system’s diagnosis module uses the ANFIS classifier to categorize abnormal foetus images into ‘‘Early’’ cases and ‘‘Advanced’’ situations. This approach uses CNN classification methodology to identify and segment the NT region. This methodology divides each pixel in a foetal picture into normal and abnormal pixels based on the derived feature set. Then, morphological segmentation method segments the region belonging to NT in classified abnormal fetus image