Investigation of computational methods for medical image segmentation

Abstract

In digital world, Image Segmentation is a process of fragmenting a digital image into set of regions based on certain properties such as gray scale levels, texture or color to extract the requisite meaningful and useful information from a digital image. Segmentation is the key process and captious step in image interpretation and analysis. The applications of image segmentation are diversely in many fields such as Military Surveillance, object detection, fingerprint and iris recognition, pattern recognition, object-based image compression, image retrieval, image enhancement and medical image processing. In today s world, Image segmentation techniques are primarily used for prognosis and diagnosis of various In digital world, Image Segmentation is a process of fragmenting a digital image into set of regions based on certain properties such as gray scale levels, texture or color to extract the requisite meaningful and useful information from a digital image. Segmentation is the key process and captious step in image interpretation and analysis. The applications of image segmentation are diversely in many fields such as Military Surveillance, object detection, fingerprint and iris recognition, pattern recognition, object-based image compression, image retrieval, image enhancement and medical image processing. In today s world, Image segmentation techniques are primarily used for prognosis and diagnosis of various diseases like liver cancer, brain tumour, lung nodules, lung cancer and Diabetic Retinopathy because of the development of several precise and accurate segmentation methods for medical images. Automated segmentation of medical images is indispensable to assist the doctors since manual investigation leads to inter-observer variability. The prime motive of this research work is to improve the health care by early detection of severe diseases like lung cancer, Diabetic Retinopathy and Alzheimer by proposing an efficient image segmentation method and classification method. In this research for image analysis computed tomography lung image, human retinal eye fundus image and magnetic resonance imaging of brain have been taken. These input images are pre-processed by a Median filter and Adaptive Histogram Equalization. Reduction of speckle noise is achieved by median filtering. Contrast level enhancement is carried out by Adaptive Histogram Equalization. The pre-processed medical images are then segmented by various segmentation techniques and the proposed method. like liver cancer, brain tumour, lung nodules, lung cancer and Diabetic Retinopathy because of the development of several precise and accurate segmentation methods for medical images. Automated segmentation of medical images is indispensable to assist the doctors since manual investigation leads to inter-observer variability. The prime motive of this research work is to improve the health care by early detection of severe diseases like lung cancer, Diabetic Retinopathy and Alzheimer by proposing an efficient image segmentation method and classification method. In this research for image analysis computed tomography lung image, human retinal eye fundus image and magnetic resonance imaging of brain have been taken. These input images are pre-processed by a Median filter and Adaptive Histogram Equalization. Reduction of speckle noise is achieved by median filtering. Contrast level enhancement is carried out by Adaptive Histogram Equalization. The pre-processed medical images are then segmented by various segmentation techniques and the proposed method. newline