An improved computer vision method for detecting white blood cells

TL;DR

This paper introduces a novel computer vision algorithm using Differential Evolution for automatic detection of white blood cells in complex medical images, improving accuracy and robustness in ellipsoid recognition.

Contribution

It presents a multi-ellipse detection method based on optimization, specifically tailored for WBC detection in cluttered smear images, which is a new approach in this context.

Findings

High detection accuracy demonstrated on complex images

Robustness against clutter and image variability

Effective optimization-based ellipse fitting

Abstract

The automatic detection of White Blood Cells (WBC) still remains as an unsolved issue in medical imaging. The analysis of WBC images has engaged researchers from fields of medicine and computer vision alike. Since WBC can be approximated by an ellipsoid form, an ellipse detector algorithm may be successfully applied in order to recognize them. This paper presents an algorithm for the automatic detection of WBC embedded into complicated and cluttered smear images that considers the complete process as a multi-ellipse detection problem. The approach, based on the Differential Evolution (DE) algorithm, transforms the detection task into an optimization problem where individuals emulate candidate ellipses. An objective function evaluates if such candidate ellipses are really present in the edge image of the smear. Guided by the values of such function, the set of encoded candidate ellipses (individuals) are evolved using the DE algorithm so that they can fit into the WBC enclosed within the edge-only map of the image. Experimental results from white blood cell images with a varying range of complexity are included to validate the efficiency of the proposed technique in terms of accuracy and robustness.