Purchase this article with an account.
Boyu Gu, Jing Lu, Xiaolin Wang, Yuhua Zhang; In vivo measurement of retinal capillary blood flow in human eye with high-speed adaptive optics near-confocal imaging. Invest. Ophthalmol. Vis. Sci. 2018;59(9):4636. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
High-resolution adaptive optics (AO) retinal imaging offers a promising method for noninvasive study of microcirculation in human eye. We develop technical strategies for accurate measurement of blood flow velocity in human retinal capillaries.
A digital micro-mirror device based AO near-confocal ophthalmoscope (AONCO) was designed with an anamorphic imaging mechanism, which significantly increased imaging light collection efficiency and enabled high speed image acquisition up to 800 frames/second (FPS) in the living human eye. Spatiotemporal plots of the blood flow in the capillaries were generated from the successive retinal images. Custom software based upon the Radon transform was developed to measure instantaneous velocity of the blood flow from the spatiotemporal plots. Imaging strategy for rendering red blood cell (RBC) flow with optimal contrast in retinal capillaries using infrared light was optimized by focusing the imaging light directly on the capillaries. The impact of image acquisition speed on velocity measurement was assessed in videos recorded with frame rates at 200, 400, and 800 FPS.
With a near-infrared (NIR) light source (λ=795 nm), the AONCO produced clear images of RBC flows in normal human subjects. High spatial resolution and high temporal resolution afforded by high-speed image acquisition enabled tracing of RBC movements on spatiotemporal plots in all imaged capillaries with high contrast, and thereby generated continuous velocity waveform of the blood flow. Near-confocal imaging rendered characterized white blood cell (WBC) and RBC traces, allowing for sensitive evaluation of the impact of WBC on the blood flow. Examination of velocity measurements under different frame rates confirmed excellent agreement between the velocity waveforms obtained at 800 FPS and 400 FPS (correlation coefficients > 92.7%). It also indicated that the velocity assessed in images acquired at 200 FPS showed significant alias in comparison to that measured at 800 FPS.
We have demonstrated that the AONCO can directly image RBC flow in most macular capillaries in the living human eye with a NIR light source. The method has a great potential to facilitate in vivo study of retinal microcirculation and pathophysiology.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only