Abstract
Purpose :
To test UC Davis 2nd generation multimodal retinal imaging system equipped with tracking scanning laser ophthalmoscope (TSLO) to stabilize the optical coherence tomography angiography (OCTA) acquisition in real-time. In parallel, ocular aberrations are recorded with a custom Shack–Hartmann wavefront sensor (SHWS). The system has two key applications: production of stabilized OCT angiograms of retina and choroidal microvasculature and screening of patients for adaptive optics (AO) imaging, based on OCT image quality, fixation stability, and aberration severity.
Methods :
The system schematic is shown in Fig. 1. All three subsystems (OCTA, TSLO and SHWS) are optically coupled with a use of dichroic mirrors before entering the eye. The wavefront detection path is then separated using a beams splitter as the ophthalmic lens is translated to correct for the patient's refractive error and thus would move the pupil plane. The TSLO uses 840 nm light to image the retina at 30 Hz frame rate with a 5° field-of-view with 960 Hz bandwidth for motion detection. The OCTA uses a 1060 nm swept laser operating at 100 kHz sweep frequency to image the retina and the wavefront sensing is done with a 755 nm superluminescent diode. The real-time motion correction from TSLO to the OCTA is done by combining voltage signals controlling the scanning mirrors and the TSLO-generated correction (motion) signal using an analog summing amplifier.
Results :
The proposed system has been tested on several subjects without known retinal disease. OCT images and angiograms have been collected with and without active tracking, and the performance of tracking system has been validated using galvanometer-based moving model eye. Data from all three subsystems have been collected from subjects who were later imaged using AO-OCT and data from the wavefront sensor has been used to exclude some patients from AO-OCT imaging.
Conclusions :
Real-time eye tracking provides two-fold benefit for the combined system. First, as OCT angiograms are generated from the volumes, eye motion produces OCT artifacts due to spurious decorrelation, which are suppressed by real-time tracking. Second, OCTA with motion tracking further improves visualization by permitting averaging of images without computational motion correction. In addition, subjects’ wavefront error can be quantified to determine whether it is suitable for the dynamic range of the sensors and correctors in the AO systems.
This is a 2021 ARVO Annual Meeting abstract.