Quantitative mapping of hemodynamics in the lung, brain, and dorsal window chamber-grown tumors using a novel, automated algorithm.

OBJECTIVE: Hemodynamic properties of vascular beds are of great interest in a variety of clinical and laboratory settings. However, there presently exists no automated, accurate, technically simple method for generating blood velocity maps of complex microvessel networks. METHODS: Here, we present a novel algorithm that addresses the problem of acquiring quantitative maps by applying pixel-by-pixel cross-correlation to video data. Temporal signals at every spatial coordinate are compared with signals at neighboring points, generating a series of correlation maps from which speed and direction are calculated. User-assisted definition of vessel geometries is not required, and sequential data are analyzed automatically, without user bias. RESULTS: Velocity measurements were validated against the dual-slit method and against in vitro capillary flow with known velocities. The algorithm was tested in three different biological models in order to demonstrate its versatility. CONCLUSIONS: The hemodynamic maps presented here demonstrate an accurate, quantitative method of analyzing dynamic vascular systems.

Duke Scholars

Author Michael Martin Haglund Neurosurgery

Author Gregory M. Palmer Radiation Oncology

Author Mark Wesley Dewhirst Radiation Oncology