Chapter 3: Intelligent Systems

Title: A mobile enabled intelligent scheme to identify blood cancer for remote areas - cell membrane segmentation using marker controlled watershed segmentation phase

Author(s): W. Srisukkham, P. Lepcha, M.A. Hossain, L. Zhang, R. Jiang, H.N. Lim

Address: Department of Computer Science, Chiang Mai University, Thailand | Computational Intelligence Group, Northumbria University, UK | Computational Intelligence Group, Northumbria University, UK | Computational Intelligence Group, Northumbria University, UK | Computational Intelligence Group, Northumbria University, UK

Reference: Software, Knowledge, Information Management and Applications (SKIMA 2013) pp. 91 - 98

Abstract/Summary: This research aims at developing an intelligent mobile (with microscopic lens) enabled decision support scheme to aid the diagnosis of blood cancer (leukemia) at an early stage so that the disease can be managed better. The first step of diagnosis is the segmentation and study of morphological characteristics of white blood cells (WBCs). Cells segmentation is a challenging problem due to the complexity and nature of the blood images. Traditional methods with human involvement make the segmentation process slow, error prone and costly. In this paper, we demonstrated the segmentation of the WBCs from a noisy background with touching red blood cells (RBCs) using modified marker controlled watershed segmentation with the integration of image processing operations. The accuracy of the segmentation achieved was 91.33% and 85.33% with and without the use of filters respectively, clearly demonstrates the potential to be used as a preliminary step for the diagnosis of leukaemia.

Order a copy of this article