Indian Journal of Science and Technology
Year: 2017, Volume: 10, Issue: 45, Pages: 1-10
Daizy Deb and Sudipta Roy
Department of Computer Science and Engineering, Assam University, Silchar, Assam − 788011, India; [email protected], [email protected]
Objectives: In this paper, we focus our research on detecting brain tumours for various diagnostic purposes in medical field. Methods/Statistical Analysis: The method applied here is soft computing for image segmentationto detect the brain tumour from a particular MRI image which is important for various diagnostic purposes in medical field. For the purpose of identifying abnormal cells in the MRI images, which are collected from various real time situations, have been passed through a de-noising algorithm followed by a clustering approach. Findings: A new Fuzzy C Means clustering method followed by multilevel thresholding and level set algorithm have been adopted to recognise the tumour affected areas. This method has been compared against existing two techniques like multilevel thresholding and K-means algorithm. K-means algorithm is more efficient regarding time but this improved technique of image segmentation ensures more precise result. Application/improvements: This algorithm is fully tested with various medical images like MRI images and also working nicely to achieve orientation of accurate shape and size of brain tumor.
Keywords: Erosion, Fuzzy C-Mean (FCM), Image Segmentation, K-Means, Level Set, Multilevel Thresholding
Subscribe now for latest articles and news.