Segmentation and classification of brain tumors using hierarchical topology preserving map

Abstract

Segmentation of an image is the separation or division of the image into different regions of similar feature In medical field Magnetic Resonance Image is used to distinguish pathological tissues from normal tissues especially for brain tumors These days millions of medical images have been produced routinely in medical care centers Usually analysis of this lump amount of data has been performed manually Even experienced and trained newlineradiologists find difficulty in analyzing a very small amount of images So physicians and radiologists are aided with computer based methods for diagnosing the patients This is the beginning of the research to create a vast database for medical images In Computer Aided Systems the analyzed computer based output has been used as a second opinion for physicians and radiologists to analyze and diagnose the patient details in a faster manner as compared to manual process Using the automated CAS identification of different tissues and pathologies is clear accurate and more certain In this research work the development of a method to assist the medical experts in the process of segmenting brain tumors using Hierarchical Topology Preserving Map is proposed The main objective is to develop a system that can follow a medical technician s way of work newlineconsidering his experience and knowledge More concretely a fully automatic and unsupervised segmentation method which considers human knowledge is presented The method successfully manages the ambiguity of Magnetic Resonance image features being capable of describing knowledge about the tumors in vague terms newline newline