Abstract
Purpose :
Quantification of retinal blood flow shows promise as an early biomarker for diverse ocular pathologies such as diabetic retinopathy and glaucoma. We have developed a non-invasive retinal imaging device—the XyCAM RITM—that captures high-resolution fundus images and uses laser speckle contrast imaging (LSCI) to estimate retinal blood flow.
Methods :
The right eye of 4 healthy individuals (58-59 years old, 2 male) received tropicamide for pupil dilation and was imaged using the investigational XyCAM RI in a clinical environment. Subjects’ heads were stabilized on a standard clinical chin rest and the XyCAM RI mounted on a slit lamp base for fine motion adjustment (Fig. 1A). Four imaging sessions were conducted on each subject, where each session comprised of an ophthalmologist focusing on a desired region of interest (ROI) in the subject’s retina, followed by rapid, automatic acquisition of 5 red-free fundus photographs (with green illumination) and 120 speckle images (using red laser illumination). For reference, color fundus photographs were also obtained using a Topcon TRC 50DX fundus camera.
At least 4 vessel ROIs were identified in each fundus photograph and retinal blood flow velocities (RBFV) were estimated within those ROIs using LSCI-based methods. For each subject, statistical analyses were completed using data from the session with the lowest variation in RBFV estimates.
Results :
Among all 4 subjects combined, the analyzed ROIs had a mean RBFV of 92.03 ± 40.47. The intra-session coefficient of variation (CV) in RBFV measurements was found to be low across subjects [CV = 6.31% ± 2.40%]. The vessel-to-background contrast ratio and variability in the intensity profile of images were further found to be similar between subjects’ XyCAM-based fundus photographs and reference images [p’s > 0.30].
Conclusions :
Preliminary results from our early feasibility clinical study validate the ability of the XyCAM RI to not only acquire fundus images with the same level of vessel-to-background contrast as color fundus photography but also complement this information with reliable functional blood flow information (Figs. 1B, 1C). Further investigation is needed to confirm these findings and demonstrate the feasibility of using the device for early diagnosis of ocular diseases.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.