Design and implementation of an efficient level set segmentation and classification for brain MR images

Abstract

Nowadays, according to the fact the excess brain segmentation is one of the most prevalent diseases of human brain–body system. Further determination of the exact location of brain segmentation in the body is a big challenge. As of now, there seems to be no scientific tool which precisely determines the presence of brain segmentation. However, mammography has shortcomings which yields 34% false negative rate which is too high. This has been overcome by digital mammography but has limitations regarding the X-ray exposure. Moreover, image cannot be altered and film processing will be slow. Brain segmentation detection rate of 7.62% can be achieved through CAD-based techniques. Although image segmentation in CAD-based method has its advantage over spatial intensity, challenge is estimating the proper prior distribution. There are numerous medical imaging methods, viz., magnetic resonance imaging (MRI), X-ray computed tomography (CT), ultrasound imaging (US), etc., that can examine different factors of human body. The detection of brain segmentation is crucial for the doctor in order to determine the status of the brain segmentations and to visualize any abnormalities that are present in the brain segmentation. The detection of anomalies of brain segmentation inside the body is a topmost field of study in medical research using biomedical image processing. Certain defects (speckle noise) in ultrasound or MRI images or US or CT and artifacts result in wrong diagnosis that could happen by analysing the scanned image. Consequently, in this proposed work the key focus is to design the algorithm based on level set segmentation, wavelets filters and artificial neural network (ANN) architecture for real-time detection of brain segmentation using biomedical images with the help of MATLAB with maximum accuracy of 94.8%.