Abstract
Purpose :
To increase the spatial resolution of oximetric measurement in small retinal microvessel vessels from about 75µm, down to ~30µm, and to carry out automatic quantitative analysis of the data acquired by the (RFI).
Methods :
Data analysis by the new software was carried out on eyes of 16 healthy and 21 diseased subjects. It was acquired with the RFI according to standard Helsinki approved protocol. Retinal images were acquired with a FOV of 20° filed (4.3x4.3 mm2) or 35° (7.4x7.4mm2). The RFI includes a fast stroboscopic flash system to acquire short movies and a fast filter wheel to change the color of each flash. The quality of quantitative oximetric analysis is notoriously difficult particularly from small vessels because of signal to noise ratio limitations. To reduce the noise we collected multiple series of 8 flashes and in each series we collected 8 images at 4 different wavelength of 555, 569, 577 and 586nm, using the rapid filter wheel. This facilitated having three sets for ODR calculations rather than one customarily done by OxyMap. We averaged their independent results. To further reduce the noise novel algorithms were developed.
Results :
We completed the algorithms and the program for the fully automatic quantitative analysis thus replacing the previous RFI’s semi-automatic but only qualitative analysis. The new software generates a topological map of the vessels and micro vessels; (ii) Measures the diameter at every position along each vessel in the images (iii) Measures the net vessel absorption. We validated the present results by comparing them to oximetry maps publications. We found that most published data included oxygen saturation values only down to vessels diameter of ~75µm. Even in larger vessels oxygen saturation values appeared noisy. We determined oxygen saturation values down to 30±7µm using a FOV of 20°. Average diameter was 46.5µm. Range 12-180µm. The relative STD of oxygen saturation values for the three ODR pairs (555,569; 569,577; 577,586), for all the pixels along the vessels was 15.3%.
Conclusions :
Mapping oximetry at high spatial resolution should prove useful for early diagnosis, preventive treatment and its monitoring. The RFI system and its new fully automatic analysis software now provide multi qualitative and quantitative measurements of functional and structural parameters.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.