Title: A new pointing kernel classifier for MRI brain tumour detection and classification

Authors: R. Meenakshi

Addresses: Department of Information Technology, Saveetha Engineering College, Thandalam, Chennai, Tamil Nadu 602105, India

Abstract: In medical imaging, detecting and classifying the brain tumours in Magnetic Resonance Image (MRI) is a demanding and critical task. This paper proposes a new system for MRI brain tumour segmentation and classification. The Distribution based Adaptive Filtering (DAF) technique is developed to remove the speckle and white Gaussian noise in the given image. In segmentation stage, the clustering and label formation processes are performed to predict the tumour part. The Neighbouring Cellular Automata (NCA) model is proposed for clustering. The labels such as Back Ground (BG), border area, Gray Matter (GM) and White Matter (WM) are formed for the clustered image. The features of the segmented image are extracted by using the Differential Binary Pattern (DBP) technique. The firefly optimisation technique is employed to select the best features. The Pointing Kernel Classifier (PKC) is used to classify the abnormal and normal brain images and the type of brain tumour.

Keywords: brain tumour; benign; malignant; metastatic; DAF; distribution based adaptive filtering; MRI; magnetic resonance image; DBP; differential binary pattern; gray matter; white matter; PKC; pointing kernel classifier.

DOI: 10.1504/IJBET.2017.085142

International Journal of Biomedical Engineering and Technology, 2017 Vol.24 No.3, pp.237 - 263

Received: 17 Feb 2016
Accepted: 25 May 2016
Published online: 13 Jul 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article