Performance Evaluation of Genetic Algorithm for Medical Image Retrieval

Abstract

Medical imaging is the procedure of generating diagnostic and treatment images of the human body, utilizing techniques such as traditional x-rays, magnetic resonance imaging, and positron emission tomography. The sheer volume of digital medical images stored in repositories poses a significant challenge for efficient access. To address this, content-based image retrieval is employed on visually analyzing the contents of a query image to retrieve relevant images. CBIR involves two key processes: feature extraction and feature matching. The primary hurdle in CBIR lies in developing flexible methodologies capable of processing diverse images with varying characteristics like color, shape, etc. This chapter concentrates on the retrieval of medical imagery as of various data sources and assesses the performance of machine learning classifiers, including support vector machine methodologies, with the aim of enhancing classification accuracy. In order to enhance the performance of classifiers, genetic algorithm is used as the tool for optimizing and for decision making for quick retrieval.