Automated ovarian masses extraction in CT images based on division of image

Abstract EN

Medical image processing is the most challenging field now a days.

Processing of CT ovarian images is one of the part of this field.The goal of this work is to present a method for detection of ovarian mass from Computed Tomography Image and determine location as well as calculation area of it.

Preprocessing of the CT image includes image resizing, conversion to gray and enhancement makes it ready for applying the processing phase which applied operations on processed image by histogram and marker controlled watershed segmentation to segment it to a set of segment that collectively cover the entire image ,then apply morphological operations to visualize only parts of masses in CT images.

Because of the nature of ovarian CT images which contain left and right lobes, a proposed algorithm to division the images after extraction the masses into two images and then apply algorithm to calculate the area of mass in each parts of image.

when execute the proposed method, the results obtained are good compared with calculation area of the mass of the ovarian extracted by the region of interest algorithm (ROI) using cursor.

A detailed procedure using Matlab (R2010a) software is written to extract mass region in CT Scan ovarian Image.