We present a non-linear 2-D/2-D affine registration technique for MR and CT modality images of section of human brain. Automatic registration is achieved by maximization of a similarity metric, which is the correlation function of two images. The proposed method has been implemented by choosing a realistic, practical transformation and optimization techniques. Correlation-based similarity metric should be maximal when two images are perfectly aligned. Since similarity metric is a non-convex function and contains many local optima, choice of search strategy for optimization is important in registration problem. Many optimization schemes are existing, most of which are local and require a starting point. In present study we have implemented genetic algorithm and particle swarm optimization technique to overcome this problem. A comparative study shows the superiority and robustness of swarm methodology over genetic approach.

No abstract

Biomedical image registration or geometrical alignment of 2-D/2-D image data is increasingly important in diagnosis, treatment planning, computer-guided therapies and in biomedical research. In this paper we present affine registration of same modality images and different (MR & CT) modality images. Automatic registration is achieved by maximization of a similarity metric, which is Mutual Information (MI) or Relative Entropy, based on the concept of information theory. Registration based on MI usually requires an optimization technique to achieve correctly aligned images. There exist many optimization schemes, most of which are local and require a starting point. Unfortunately the functions of similarity metric used in the present problem are nonconvex and irregular and therefore global methods are often required. In this paper, we have implemented Genetic algorithm as an optimization technique to overcome these problems. Experimental results show our algorithm is a robust and efficient method.

Abstract:The problem of feature selection consists of finding a significant feature subset of input training as well as test patterns that enable to describe all information required to classify a particular pattern. In present paper we focus in this particular problem which plays a key role in machine learning problems. In fact, before building a model for feature selection, our goal is to identify and to reject the features that degrade the classification performance of a classifier. This is especially true when the available input feature space is very large, and need exists to develop an efficient searching algorithm to combine these features spaces to a few significant one which are capable to represent that particular class. Presently, authors have described two approaches for combining the large feature spaces to efficient numbers using Genetic Algorithm and Fuzzy Clustering techniques. Finally the classification of patterns has been achieved using adaptive neuro-fuzzy techniques. The aim of entire work is to implement the recognition scheme for classification of tumor lesions appearing in human brain as space occupying lesions identified by CT and MR images. A part of the work has been presented in this paper. The proposed model indicates a promising direction for adaptation in a changing environment.

In the present work, authors have developed a treatment planning system implementing genetic based neuro-fuzzy approaches for accurate analysis of shape and margin of tumor masses appearing in breast using digital mammogram. It is obvious that a complicated structure invites the problem of over learning and misclassification. In proposed methodology, genetic algorithm (GA) has been used for searching of effective input feature vectors combined with adaptive neuro-fuzzy model for final classification of different boundaries of tumor masses. The study involves 200 digitized mammograms from MIAS and other databases and has shown 86% correct classification rate.