Automatic Detection of White Matter and Gray Matterfrom Mri Image for Diagnosis of Neurological Diseases
Year: 2020
Type: article
Abstract: The brain is the most complex organ which acts as the centre of the nervous system of the human body.
Quantitative analysis of Gray Matter (GM) and White Matter (WM) is vital part for detection of multiple brain
disease such as asymptomatic unruptured aneurysms, Alzheimer's disease, vascular dementia, and multiple sclerosis
(MS). MRI is very useful medical imaging technique which helps the doctor for diagnosis and therapy of these
kinds of complex neurological diseases. Quantification of white matter lesions is necessary for drug treatment
assessment multiple sclerosis, while in schizophrenia and epilepsy, volumetric of gray matter, white matter, and
cerebrospinal fluid is important to characterize morphological differences between subjects. Manual segmentation is
too time consuming since vast amounts of data are required for such studies. Moreover, such manual segmentations
show large inter and intra observer variability. Hence, there is a need for automated segmentation tools. This paper
proposes an efficient algorithm for automatic segmentation of brain tissues such as gray matter and white matter
regions from MR images. An efficient algorithm has been proposed for skull stripping. Hybrid clustering method
which is combination of k-means and Fuzzy C-Means (FCM) has been achieved optimal result. Intensity values and
statistical feature based values are used to measure the accuracy of our proposed hybrid segmentation algorithm.
Final result of clustering gray and white matter compared with the manually marked ground truth using standard
accuracy measurement coefficients. The performance of the statistical feature clustering is more promising and
produce prominent result compare to intensity based clustering. In this study experimental result and validation
indicate the accurate detection and segmentation from MRI brain image.
Source: Journal of Critical Reviews
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W3204477607