Abstract
Purpose:
We evaluated a commercially available flood-illuminated adaptive optics (AO) camera for its reliability and efficacy to provide structural diagnostic information in subjects with inherited retinal degenerative diseases.
Methods:
We used the RTX1 flood-illuminated adaptive optics camera (Imagine Eyes: Orsay, France) to image 105 subjects with inherited retinal degeneration ranging in age from 8 to 88 years old. For each subject, a series of 25 4°x4° retinal images were obtained in one or both eyes. Using i2k Retina (DualAlign LLC, Clifton Park, NY, USA), these images were combined to create a retinal montage spanning a 12°x12° field of the central macula. In a subset of higher quality images, cone counting was performed automatically by applying background subtraction and thresholding of local maxima in Matlab (MathWorks, Natick, MA, USA). All testing was approved by the OHSU IRB.
Results:
Image quality varied substantially between individual subjects and across pathologies. If the 9 central tiles and at least 20 of the 25 total images could be used to form a montage, then the imaging session was considered successful. Twenty-six out of 40 eyes (65%) with retinitis pigmentosa (RP), 9 out of 15 eyes (60%) with Usher syndrome, and 7 out of 12 eyes (58%) with Stargardt dystrophy were successfully imaged. Thirty-five out of 80 eyes (44%) with other less commonly tested pathologies resulted in successful imaging. The ability to see cone photoreceptors by AO usually correlated to the visualization of the ellipsoid zone by optical coherence tomography (OCT). In areas where the ellipsoid zone appeared severely damaged when imaged by OCT, cone-like hyper-reflective spots were often seen with AO. An annulus of blur was consistently seen at varying eccentricities in RP patients and was correlated to autofluorescence and OCT images. Lens artifacts were consistently seen in the AO images of subjects that had received cataract surgery.
Conclusions:
Current flood illuminated adaptive optics technology is most effective when imaging subjects with good central vision, such as RP and Usher patients. Additionally, caution should be used when interpreting flood-illuminated AO images due to the fact that it is a relatively new and under-studied retinal imaging technique.
Keywords: 550 imaging/image analysis: clinical •
688 retina •
465 clinical (human) or epidemiologic studies: systems/equipment/techniques