Synthetic images of blood microcirculation to assess precision of velocity profiles by a cross-correlation method

Optical cross-correlation methods have been used to study the motion of red blood cells (RBC) in the microcirculation. To evaluate the precision of such a method to determine RBC velocity profiles, we developped a computational model of the microscopy image formation. The following steps were undertaken: (1) a mechanical model was used to mimic three dimensional RBC movements in a tubular parabolic flow; (2) at each time step, a synthetic image was built using microscopic image formation equations based on the depth of correlation of RBCs; and (3) the velocity profile was extracted by a cross-correlation algorithm applied to these synthetic images. The estimated maximum velocities extracted from the simulated images were always smaller than velocities found by simulation. Relative errors (4% to 25%) depended on the vessel radius and on the shape of the velocity profile, but not on the hematocrit or on the maximum velocity.