Abstract
Purpose: :
To visualize in 3D and quantify in-vivo outer retinal abnormalities in the rat retina induced by a retinal toxin (NaIO3), by use of a high speed, ultrahigh resolution optical coherence tomography (UHROCT) system and novel image processing algorithms.
Methods: :
A high speed, UHROCT system, operating in the 1µm wavelength range was developed for non-invasive rodent retina imaging. The system provides 3µm axial and ~5µm lateral resolution in the rat retina and 47,000 A-scans/sec image acquisition rate. Novel speckle denoising and automatic segmentation algorithms were developed and applied to the UHROCT tomograms for improving the visualization of fine morphological detail in the retina, and for automatic evaluation of the thickness of individual retinal layers. RPE and outer retinal damage was induced by intravenous injection of sodium iodate (NaIO3) in adult, female Long Evans rats. 3D UHROCT tomograms were acquired at baseline and at days 3 and 7 of the longitudinal study. At the end of testing, eyes were removed and processed by Haemotoxylin & Eosin (H&E).
Results: :
2D and 3D OCT images of healthy rodent retinas showed clear visualization of all retinal layers, RPE, and choroidal vasculature. UHROCT tomograms acquired 3 and 7 days post NaIO3 injection showed progressive loss of the RPE layer, morphological changes in the inner- and outer segments of the photoreceptor layer appearing as clusters of highly reflective matter of ~10µm - 50µm in diameter and alternating with areas of abnormally low signal in the RPE, complete loss of the ELM as well as changes in the shape and thickness of the outer nuclear layer. Histological analysis confirms the loss of RPE, swelling and folding of the outer retina, loss of photoreceptor outer/inner segments, along with in some areas, complete loss of the outer retina such that the inner retina is in direct contact with Bruch’s membrane.
Conclusions: :
We have developed a state-of-the-art, UHROCT system for imaging the rodent eye and image processing algorithms for automatic segmentation and thickness measurement of retinal layers. When applied to a retinal toxicity model in the rat eye, the system allows for in-vivo observation and precise quantification of morphological changes that correlate well with histology.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • retina • degenerations/dystrophies