Model experiments on measuring flow in microvessels using tracers.

Most techniques for measuring plasma or red cell flow velocity within microvessels rely on determining the transit time of a tracer to transverse the distance between two monitoring sites within a vessel. In principle, proper transit time determinations require flow-weighted sampling of the tracer at monitoring sites. In practical application of the tracer technique, however, trace sampling at monitoring sites is not flow-weighted but is area-weighted, and hence elapsed transient time can only be estimated from tracer data. We previously showed theoretically (Microvasc. Res. 40, 394-411, 1990) that the flow velocity determined under these conditions can differ appreciably from the actual mean flow velocity of the carrier fluid within the microvessel. Nevertheless, trace mean flow velocity does approach that of the fluid when tracer velocity is measured past a finite distance from the microvessel entrance. In this study, we examined the tracer measurement of flow experimentally using a physical model. We perfused single glass microvessels and simple fabricated microvessel networks with distilled water at physiological flow rates. Mean tracer velocity (Vd) was determined at several axial locations within the microvessels using injected Evans blue dye. At each location Vd was determined in a manner consistent with usual application of the tracer flow measurement technique. Actual mean flow velocity (Va) was determined from the measured effluent flow rates discharged from each microvessel. Our experimental results confirm the existence of an appreciable velocity measurement error (VME) associated with the tracer technique. The VME behavior was consistent with our original theoretical analysis. Vd was significantly smaller than Va within a finite length of vessel near the entrance, but approached and became equal to Va past this length. Furthermore, even under conditions where the VME was negligible at the end of a parent microvessel, a new and appreciable VME arose within the connecting downstream daughter microvessels. These results underscore that tracer mean flow velocity as obtained in normal implementation of the tracer technique can differ from that for the fluid itself within a microvascular network.