Segmentation and detection of white blood cells in images using artifitial inmune systems

Authors

  • Erick Cuevas Universidad de Guadalajara, CUCEI, Departamento de Ciencias Computacionales.
  • Valentín Osuna Enciso Universidad de Guadalajara, CUCEI, Departamento de Ciencias Computacionales
  • Diego Oliva Universidad de Guadalajara, CUCEI, Departamento de Ciencias Computacionales
  • Fernando Wario Universidad de Guadalajara, CUCEI, Departamento de Ciencias Computacionales
  • Daniel Zaldivar Universidad de Guadalajara, CUCEI, Departamento de Ciencias Computacionales

Abstract

 

White blood cells also known as leukocytes play a significant role in the diagnosis of different diseases. Digital image processing techniques have successfully contributed to analyze the cells, leading to more accurate and reliable systems for disease diagnosis. However, a high variability on cell shape, size, edge and localization complicates the data extraction process. Moreover, the contrast between cell boundaries and the image’s background varies according to lighting conditions during the capturing process. On the other hand, Artificial Immune Segmentación Systems (AIS) have been successfully applied to tackle numerous challenging optimization problems with remarkable performance in comparison to classical techniques. One of the most widely employed AIS approaches is the Clonal selection algorithm (CSA) which allows to proliferate candidate solutions (antibody) that better resolve a determinate objective function (antigen). This paper is thus focused on the segmentation, localization and measurement of leukocytes from other different components in blood smear images. The algorithm uses two different CSA systems, one for the segmentation phase and other for the leukocyte detection task. In the approach, the detection problem is considered to be similar to an optimization problem between the leukocyte and its best matched circle shape. On the other hand, Clonal selection algorithm (CSA) is one of the most widely employed AIS approaches. The algorithm uses the encoding of three points as candidate leukocytes over the edge image. An objective function evaluates if such candidate leukocytes are actually present in the edge image following the guidance of the objective function’s values. A set of encoded candidate leukocytes are evolved using the CSA so that they can fit into the actual leucocytes shown by the edge map of the image. The obtained results in comparison with the stateof-the-art algorithms validate the efficiency of the proposed technique with regard to accuracy, speed, and robustness.

Downloads

Download data is not yet available.

Published

2010-11-03

How to Cite

Cuevas, E., Osuna Enciso, V., Oliva, D., Wario, F., & Zaldivar, D. (2010). Segmentation and detection of white blood cells in images using artifitial inmune systems. Revista Mexicana De Ingenieria Biomedica, 31(2), 16. Retrieved from https://rmib.mx/index.php/rmib/article/view/448

Issue

Section

Research Articles

Dimensions Citation