Robotically aligned OCT enables physically distanced imaging and relevant measurements in a retina clinic population

Abstract

Purpose : Optical coherence tomography (OCT) is a ubiquitous ophthalmic imaging technology. However, currently patients are required to position themselves in chin/forehead rests for stabilization with the system operator in close proximity. Introducing a robot to the clinic enabled distanced, no-contact imaging of a retina clinic population.

Methods : We developed a custom robotically (UR3e) aligning swept source OCT (RAOCT; 1040nm) system with a low mass (2.4kg) sample arm utilizing 70mm diameter achromatic lenses with 3D printed optomechanics providing a clinically relevant 32° view on the retina (Fig. 1A&B). During acquisition, face and pupil tracking cameras triangulated 3D eye and pupil motion which were compensated in real-time by the robot and OCT system (Fig. 1C&D). We acquired B-scans and volumes in triplicate from the right eyes of 20 subjects from the Duke Eye Center clinics (10 normal, 10 diseased; 25 - 91 years old) with RAOCT and Heidelberg Spectralis under an IRB approved protocol. Differences between device foveal thickness maps were tested using two-tailed t-tests.

Results : Subjects were seated and free of any head restraints (no forehead strap or chin rest, Fig. 1A). The system automatically aligned on the subject’s eye allowing for motion compensated OCT B-scans and volumes of the retina (Fig. 2). There was a mean paired inter-device 1.2 ± 5.9µm difference (p = 0.53) in healthy retinal thickness and 3.7 ± 7.5µm (p = 0.13) difference in diseased retinal thickness (Fig. 2E).

Conclusions : In a clinic population, we demonstrated a robotically aligned OCT system that provides retinal views and measurements comparable to current clinical OCT.

This is a 2021 ARVO Annual Meeting abstract.

 

A) RAOCT system at Duke Eye Center. Operator 2m from subject, behind Plexiglas barrier. Red arrow, face tracking camera. B) RAOCT optical system schematic C) 3D eye position estimation over OCT acquisition time covering 1.91mm range of motion. D) Resultant RAOCT motion compensated volume acquired during (C). White arrow, fovea; Yellow arrow, optic nerve head.

 

A) Averaged RAOCT B-scan of healthy retina B) Retinal thickness maps from volunteer in (A) C) Averaged RAOCT B-scan from patient with multifocal choroiditis and panuveitis D) Retinal thickness maps from volunteer in (C) E) Boxplot of measured foveal thicknesses.