Title: A classification of chronic leukaemia using new extension of k-means clustering and EFMM based on digital microscopic blood images

Authors: C. Kalaiselvi; R. Asokan

Addresses: Department of Electronics and Communication Engineering, Kongu Engineering College, Erode, Tamil Nadu, India ' Department of Electronics and Communication Engineering, Kongu Nadu College of Technology, Trichy, Tamil Nadu, India

Abstract: Leukaemia is a cancer of the white blood cells. The type of white blood cell affected in either lymphoid or myeloid. And leukaemia is defined in two ways, such as acute leukaemia (AL) and chronic leukaemia (CL). These kinds of leukaemia start when typical blood cells change and grow wildly. This paper describes in the following steps to classify the chronic leukaemia automatically and more accurately. First, pre-processing the colour scale of digital microscope blood image, then segment the image by new extension of k-means clustering algorithm, and Hausdorff dimension (HD) is utilised for feature extraction, finally the classification is done by utilising Enhanced Fuzzy Min Max (EFMM) neural network. The proposed method obtained 99.95% accuracy for Lymphocytic and Myelogenous cells.

Keywords: pre-processing; k-means clustering; Hausdorff dimension; EFMM neural networks; classification; chronic leukaemia; digital images; medical images; microscopic blood images; colour scale; image segmentation; feature extraction; enhanced fuzzy min max; lymphocytic cells; myelogenous cells; white blood cells.

DOI: 10.1504/IJBET.2017.082664

International Journal of Biomedical Engineering and Technology, 2017 Vol.23 No.2/3/4, pp.232 - 241

Received: 26 Apr 2016
Accepted: 27 Jul 2016

Published online: 25 Feb 2017 *

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