Abnormality recognition and feature extraction in female pelvic ultrasound imaging

Abstract

The detection of uterine abnormalities in the early stage is challenging and it aims to play a vital role in automated screening systems. The algorithm proposed herein works in the following way: initially a PDE (Partial Differential Equation approach) based speckle reduction is carried out in a preprocessing stage, and the region of interest is separated from the female pelvic ultrasound images. Then the texture and shape features are extracted from the segmented ROI component. The results of shape similarity indices between manual and automated segmentations are also illustrated. The simulation result analysis of this algorithm was implemented in MATLAB R2016a. In this paper, an automatic segmentation scheme is proposed which will aid the radiologists for better outputs in very limited time making the proposed solution very effective and productive. Keywords: Speckle, Segmentation, Tumor, Feature extraction and ultrasound imaging