Determination of Erythrocyte Flow Velocity by Dynamic Grey Scale Measurement Using Off-Line Image Analysis.

Yuying Liu,Jiying Yang,Kai Sun,Chuanshe Wang,Jingyan Han,Fulong Liao
DOI: https://doi.org/10.3233/ch-2009-1238
2009-01-01
Clinical Hemorheology and Microcirculation
Abstract:Determination of erythrocyte flow velocity is an important parameter for microcirculatory research. We established an off-line method for determining erythrocyte velocity in microcirculation by computer- ized dynamic grey scale analysis. In principle, the method is based on the dual-slit photometric technique (1), and updated by software image analysis. We have applied high speed video recorder (FASTCAM- ultima APX, Photron, San Diego, CA) to capture digital images for mesentery microvessels in rat and rabbit since 2004. Each series of images during 10 seconds is captured at a speed of 2000 frames/s (fps) and is composed of images of 704 × 576 pixels with 256 grey levels. The video recording is converted to AVI format required by the software of ImagePro Plus (versions 4.5 or 5.0). The video is then inves- tigated frame by frame with the help of the "Sequencer Toolbar" in the menu of ImagePro Plus. The criterion for selecting an investigative segment in a capillary is that the erythrocyte flow keeps smooth and steady for 10 seconds at least. A reference line perpendicular to the axis of the capillary is setup to show the grey scale distribution along the radial line in the first frame of image. Then, a down-stream measuring line perpendicular to the axis of the capillary is setup with a distance about 10 µm apart from the reference line. The dynamic grey scale changes at the measuring line in the subsequent images are monitored frame by frame and the changes are compared with the grey scale distribution at the reference line in the first image (Figs 1 and 2). When the most matchable grey scale with the reference distribution is detected, the time interval for erythrocytes flowing from the reference line in the first frame to the measuring line in the "matchable" frame can be figured out by multiplying the frame difference by the duration between each frame. The erythrocyte velocity can be easily calculated by dividing the distance between the reference line and the measuring line by the time interval (Fig. 2).
What problem does this paper attempt to address?