Leukocyte Tracking

leukocytetracking pic.jpg

Project Info

Title Advanced Biomedical Image Processing for Tracking Leukocytes In Vivo from Video Microscopy
Goal Design and build a completely automated and digital leukocyte tracking system.
Funding National Institutes of Health (NIH) and the Whitaker Foundation


To more accurately, effectively, and efficiently collect leukocyte behavior in order to develop anti-inflammatory drugs to treat diseases such as Crohn's disease, heart disease, and multiple sclerosis.


During the inflammation process the body uses leukocytes (white blood cells) to fight infections and help repair damaged tissues. However, cases exist when the body improperly triggers the inflammatory response process causing leukocytes to attack healthy tissue. These situations can result in a range of conditions and diseases, such as arthritis and heart disease. The movement of leukocytes and their interaction with the endothelium (vessel wall) provides valuable information about the inflammation process. Advanced automated tracking algorithms provide the necessary capability to investigate leukocyte motion within living animals.

Tracking algorithms we tested include a centroid tracker, a correlation tracker, an enhanced centroid tracker, an enhanced correlation tracker and several snake trackers. All data was collected from venules in the cremaster muscle of a living mouse. Data is collected using optical and digital cameras mounted on microscopes. The centroid and correlation trackers fail because they lack the robustness an in vivo tracker requires. The enhanced centroid and enhanced correlation trackers improve on the basic centriod and correlation shortcomings by using adaptive template matching, but lack potential to be developed into a robust biomedical tracker. However, the snake tracker shows promise to be engineered into a robust biomedical tracker. Below are compressed avi files containing examples of these five types of trackers.
  • Centroid tracker [.avi]
  • Correlation tracker [.avi]
  • Enhanced centroid tracker [.avi]
  • Enhanced correlation tracker [.avi]
  • Snake tracker [.avi]

We work in conjunction with Dr. Klaus Ley form the Biomedical Engineering Department. Dr. Ley requested the development of a software program to increase the accuracy and efficiency of tracking rolling leukocytes. Our answer was the Robust Biomedical Tracking (RBT) System, delivered in July 2002. This software package has the capability to:
  • Determine vessel diameter
  • Manually track rolling leukocytes
  • Automatically track rolling leukocytes
  • Selectively track rolling leukocytes and
  • Create presentation-ready audio-video interlaced (avi) files

The RBT takes avi files or a sequence of bitmaps (bmp) as input. The RBT records the leukocyte position at every instant in time (pixels), cell diameter (microns), X and Y scale (pixels per micron), and frames per second. The RBT determines a signal-to-clutter ratio (SCR) and calculates total displacement (microns) and average velocity (microns per second). Below is a screen shot of the RBT manually tracking a leukocyte.

leukocytetracking fig1.jpg
Fig 1. RBT tracking a leukocyte.

Below is a screen shot of the RBT's Create AVI function. Click here to view a compressed avi (3.8MB) of the the Manual, Automatic, and Selective Trackers in action over 60 frames.

leukocytetracking fig2.jpg
Fig 2. RBT create AVI.

Three of the following animated gif links provide more examples of tracking rolling leukocytes. The final link provides an example of tracking actin filaments (myosin motility).
  • Rolling leukocytes in vivo [.gif]
  • Rolling leukocytes in vitro [.gif]
  • Rolling leukocytes in vivo w/ aid of fluorescent tagging [.gif]
  • Actin filaments [.gif]


Topic revision: r3 - 01 Jun 2014, AndreaVaccari
©2018 University of Virginia. Privacy Statement
Virginia Image and Video Analysis - School of Engineering and Applied Science - University of Virginia
P.O. Box 400743 - Charlottesville, VA - 22904 - E-Mail viva.uva@gmailREMOVEME.com