Title: Design and implementation of a new cooperative approach to brain tumour identification from MRI images

Authors: Boucif Beddad; Kaddour Hachemi; Sundarapandian Vaidyanathan

Addresses: LTC Laboratory, Faculty of Technology, Dr Tahar Moulay University, Saida, Algeria ' LTC Laboratory, Faculty of Technology, Dr Tahar Moulay University, Saida, Algeria ' Research and Development Centre, Vel Tech University, Avadi, Chennai 600062, Tamil Nadu, India

Abstract: Magnetic resonance imaging has become a vital component of a large number of biomedical applications and also plays a major role in medical diagnostics. In this research work, the main purpose is to carry out a new cooperative approach to brain tumour detection and identification from MRI images with good segmentation accuracy. The proposed system applies K-means algorithm to optimise the initial centroids of the improved fuzzy C-means which incorporates the spatial information and also to get a better estimation of their final cluster centres. Then the obtained results are considered as an initialisation of the active contour for level sets technique. The proposed segmentation algorithm and its improvement were well implemented practically in real-time using a floating-point TMS320C6713 DSP of Texas Instruments. Performance improvement is measured by including various optimisation techniques, and all profiling and debugging results are shown using C6713 graphical user interface.

Keywords: CCS; code composer studio; fuzzy C-mean; MRI image processing; level sets algorithm; segmentation.

DOI: 10.1504/IJCAT.2019.097113

International Journal of Computer Applications in Technology, 2019 Vol.59 No.1, pp.1 - 10

Received: 10 Oct 2017
Accepted: 21 Nov 2017
Published online: 21 Dec 2018 *

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