A novel approach to improve the Accuracy of brain tumor detection Through segmentation in mr images

Abstract

In the medical field, digital images are commonly used for diagnostic newlinepurposes. For effective treatment of tumours, early detection of brain tumours newlineis important. Due to high soft-tissue contrast and zero sensitivity to ionizing newlineradiation, Magnetic Resonance Imaging (MRI) is the most common method newlinefor diagnosing human brain tumours. The classification of brain tumours, newlinehowever, is not a trivial task. Human observation is the standard technique for newlinerecognizing and classifying MRI brain tumors, which depends heavily on the newlineexperience of radiologists examining and analyzing image characteristics. newlineMoreover, for large volumes of data, the operator-assisted classification newlinemethods are unreliable and are thus non-reproducible. It is usually time newlineintensive and difficult to manually identify brain tumors from MR imaging. newlineTo solve these problems, computer-aided diagnostic approaches are therefore newlineextremely desirable. The current study aims to create an approach for newlineautomatically segmenting brain tumor images into various pathological types. newlineIn medical field diagnostics, automatic segmentation and identification newlineof brain tumors are of significant importance as it provides information on newlinefunctional structures in addition to the potentially irregular tissue required for newlineplanning surgery. However, this is still a challenge owing to low contrast and newlinepoorly defined limitations, and accuracy issues. Thus, for automated tumor newlinesegmentation, the Refined Migrating Birds Optimization (RMBO) algorithm newlineis used, which overcomes the shortcomings of conventional metaheuristic newlinesegmentation techniques. The RMBO, including three stages, aims to enhance newlineboth migration and position upgrade measures. In the first point, with the aid newlineof an improved tracking algorithm, pre-processing and film artifacts and newlineunwanted areas (skull) of MRI images are removed newline