MRI brain cerebral cortex segmentation using HSV based Hill Climbing technique to detect Alzheimer’s Disease

• Main Authors: Rajangam, Sivanesan, Palanisamy, Kalavathi
• Format: Conference Proceeding
• Language: English
• Subjects: Algorithms; Brain; Cerebral cortex; Color; Epilepsy; Human motion; Image segmentation; Magnetic resonance imaging; Medical imaging; Motor ability; Neuroimaging; Saturation (color)
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0212525

Clinical images produced by Magnetic Resonance Image (MRI) scanners provide complete pathophysiological detail of a brain. To diagnose any kind of brain disease like a tumor, cancer, epilepsy, Parkinson’s, Alzheimer’s Disease (AD), etc., segmentation is the foremost practice for further assessment. Especially cerebral cortex (CC) segmentation is one of the toughest chores and it plays a most important part in AD detection because the CC controls the motor skills of a human. AD affects the motor skills of humans and prolonging this condition will lead to death. Therefore, we concentrated on segmenting the CC for AD detection. This paper deals with the Hue Saturation-Value (HSV) based Hill Climbing algorithm (HSV-HC) to segment the cerebral cortex from the MR brain image. The first step of the process is to pre-process the input image, the next step is employing a hill-climbing approach on the HSV color space image to obtain the local maxima of clusters in the color histogram of an image, as a result of this process the CC in the brain image are segmented. IBSR_18 and ADNI brain images are used to evaluate this proposed method based on similarity and quantitative measures. The results produced by this method are noticeable and acceptable compared to the existing standard methods.