Purpose
To develop improved image processing and analysis methods to obtain retinal oximetry images with a wide field of view and increased spatial definition.
Methods
We use the Oxymap T1, a commercial dual-wavelength, spectrophotometric retinal oximeter (Oxymap, Reykjavik, Iceland). Using a custom developed plugin for ImageJ (NIH, Bethesda, MD, USA) and commercial image registration and montaging software (i2k Retina, DualAlign LLC, USA), we obtain detailed oxygen saturation maps of arbitrarily large regions of the eye fundus by montaging several snapshots into one image. The central specular reflection on retinal vessels is automatically analyzed and eliminated using custom image processing algorithms.
Results
Our retinal oximetry images show retinal vessel oxygen saturation patterns of larger regions of the retina than was previously possible. Previous studies using the Oxymap T1 retinal oximeter have used single 50-degree field of view images centered on the optic disc to obtain a measure of the net oxygen balance of retinal arterioles and venules. Our new method combines several 50, 35 or 20-degree snapshots into an oximetry image with a wider field of view and higher spatial definition of larger areas including the optic disc, macula and more peripheral areas. Vessels supplying the macula have higher venular saturation values and lower arteriovenous difference than other parts of the retina.
Conclusions
We have developed a new method to process retinal oximetry images from the Oxymap T1 to obtain wide-field, increased spatial definition oxygen saturation maps of the retina. Increased field of view shows regional variability in oxygen levels. Increased spatial definition shows a systematic intravascular variability in oxygen saturation.