Document Type


Date of Award


Degree Name

Master of Science in Biomedical Engineering - (M.S.)


Biomedical Engineering Committee

First Advisor

Arthur B. Ritter

Second Advisor

David S. Kristol

Third Advisor

Francine Reibman


The dynamics of macromolecular transport across microvascular walls in the hamster cheek pouch have previously been studied by other investigators using fluorescence intravital microscopy. Fluorescein Isothiocyanate labeled Dextrans (FITC-Dx) of 20,000, and 70,000 molecular weight (FITC-Dx-20 and FITC- Dx-70) were used as macromolecular probes. Values of fluorescent intensity were previously recorded on videotape as a function of distance perpendicular to leaking vessels over a period of about 2.5 hr to allow for tracer equilibration in the interstitial space. In this study, these videotapes were played back frame-by-frame and 215x130-pixel windows around each leaking vessel were selected for further analysis of the leakage patterns using digital image processing. Window areas around each leaking vessel ranging in size from 1 by 1 to 10 by 10 pixels were selected. Histograms of the light intensity distributions for the fields selected at several time points were used to construct integral optical density-position profiles of the extravasated fluorochromes.

The graylevel value of a single pixel in the middle of the window was compared with the average graylevel of 5 X 5 and 10 X 10 pixel windows. We found that the larger the window area, the less the variation is in the graylevel value. We found that, although the graylevel values for single pixels were similar in magnitude to the mean values of a pixel in the 5 X 5 and 10 X 10 windows, the standard deviation of graylevel values of single pixels, were much larger than the standard deviations calculated for the mean of the windows in 10 X 10 window areas. Thus, simple spatial filtering provides a straightforward method for smoothing the intensity profiles to estimate convective and diffusive transport parameters from a "best fit" of these profiles to a partial differential mathematical model for blood-tissue exchange. In addition, spatial smoothing should produce more reliable estimates of the transport coefficients. The large variation in single pixel values within a small window can not be explained simply by experimental error. The data suggest that the spatial variation in intensity maght be explained by the heterogeneous nature of the interstitial space. Portions of the interstitium (e.g. the intercellular compartments) might be available to the macromolecular probes used in this study, whereas other portions of the interstitium (e.g. the cellular compartments) might be exclude the macromolecules used.