Computational Analysis for Segmentation of cDNA Microarray Image using Genetic Algorithm Approach

Abstract

This research work investigates an implementation for performing microarray image processing on a given DNA microarray image. The focus on image analysis and the processing of computational analysis has been explained through various methods. The methods proposed herein find application in biomedical images like: cDNA micro array images and 2D gels images obtained from the two dimensional electrophoresis of proteins. They utilize the intensity of information relating to the images and do conversion of base problems of analysis and processing inclusive of determination of the grid structure and spot segmentation for optimizing problems, and follow this up with the solution using the method of genetic algorithms. A complete program was designed and developed to do gridding, segmentation, spot location, and presentation of the results to the user in a visual manner. Spot segmentation for various DNA microarray images has been tested and validated with different microarray images with irregular spot sizes, shapes and contamination levels. Comparing with other existing programs, the developed program using Mat lab scripts is less dependent on human input, but is less flexible when locating spots for different microarray images, i.e. images that do not have a specific spot and grid layout. The proposed method result of such application have attained higher accuracy with less time in comparison with various well known techniques that find extensive application. newline