Abstract
Purpose: :
To test a new method for measuring blood flow in the human retina. This method uses an AOSLO to directly measure the time required for red blood cells (RBCs) to move across a horizontal scanning line.
Methods: :
An 840 nm (center wavelength) SLD was used as the light source. The AO control of our system was maintained using a BMC MEMS deformable mirror and a Shack Hartmann sensor operating in closed loop at 8 Hz. The imaging beam was steered by an 8 kHz horizontal scanner and a 15 Hz vertical scanner to sweep across the imaging region. The vertical scan was programmed to briefly stop on top of a blood vessel while the horizontal scan repeatedly moved the beam across that line in the imaging area. During this halt, all the image lines were collected from the same location. In this period the image consists of a direct measure of the change in time due to horizontal movement of red blood cells across the imaging region. Thus, each RBC or RBC cluster was sampled several times while passing the horizontal scanning line, forming a diagonal straight line. The slope of the diagonal line is the horizontal velocity of the red blood cell, given the constant time interval between line scans. Velocities were adjusted for the angle between the horizontal scanning line and the vessel flow. 10ms samples were obtained, avoiding contamination from eye movements. Vessels with size from 10um to 100um were sampled in healthy subjects.
Results: :
Measured velocity in both arteries and veins fluctuates with the subject’s cardiac cycle, with velocity fluctuation in arteries being larger, as expected. The peak velocity for an artery was typically 300% that of the minimum, whereas for veins there was only a 2:1 ratio in velocity across the cardiac cycle. For the arteries, the velocity waveform showed a dicrotic notch after the systolic peak, which is very typical in arterial pressure waveforms and is described in the literature to possibly result from the closure of the aortic valve. Capillary velocities measured range from 1mm/s to 2 mm/s and in larger arteries sampled velocities went as high as 40mm/s at their pulse peaks. Parabolic velocity profiles were measured across larger blood vessels. We computed the average flow rate across the cardiac cycle. The sum of the calculated flow rate in daughter vessels was very close to that in the parent vessel.
Conclusions: :
Direct imaging of blood flow is possible using an AOSLO. This technique is sensitive to known variations in blood flow in human observers. This technique has the potential to fill the gap in measurements between techniques such a Doppler velocimetry, which can measure large vessels, and leukocyte imaging which can measure small parafoveal capillaries.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • blood supply • retina