Abstract
Purpose :
Since the bulk of the oxygen is carried by the erythrocytes, in vivo characterization of the spatial and temporal distribution of erythrocytes in the retinal capillary can provide an accurate estimate of local oxygen supply and the functional status of the tissue.
Methods :
We acquired high resolution images of the erythrocytes flowing in retinal capillaries of human subjects in normal ocular and physical health, using a high-speed adaptive optics near-confocal ophthalmoscope at a frame rate of 800 Hz over a field of view 1.2° X 0.3°, and generated the spatiotemporal traces of individual erythrocytes from consecutive frames of the video. We developed a custom software using the Hough transform to extract the binarized erythrocyte traces, and thereby measured the erythrocyte lineal density (defined as the number of cells presenting within a certain length of the capillary). We evaluated the accuracy of the automated measurement by comparing results obtained using the algorithm with that obtained by manual counting. In addition, we measured the erythrocyte flux in the same segment of the capillary where we counted the cell linear density.
Results :
Results: We measured the cell linear density and cell flux in 37 macular capillaries of 10 eyes of 10 human subjects in normal ocular and physical health. The subjects were divided into 2 age groups, 20-29 years old and 50-59 years old. In the young and the old groups, the linear density was 3.20 ± 0.90 cells/50 μm and 2.97 ± 1.14 cells/50 μm, respectively. The mean difference between the results obtained with the manual measurement and the automatic algorithm was 0.0702. The 95% confidence interval was [-0.8467, 0.9871]. The cell flux was 72.74 ± 21.05 cells/s and 70.46 ± 20.79 cells/s respectively in the 2 groups.
Conclusions :
The quantification of spatial-temporal distribution of the erythrocytes in retinal capillaries in the living human eye under natural rheological state can provide the absolute values of the oxygen supply that is needed by normal retinal function. The method may facilitates the study of the fundamental physiology and pathophysiology of the microcirculation in the living human retina.
This is a 2020 ARVO Annual Meeting abstract.