Abstract
Purpose :
Optoretinography (ORG) is a tool for noninvasive functional assessment of retinal neurons. Development of stimuli and quantification of responses are critical first steps toward employing the ORG as a functional biomarker. Here we show ORG measurements from a custom adaptive optics full-field OCT system [Valente et al., 2020, BOE] using several stimulus patterns and describe a model for quantifying them.
Methods :
The system consists of a tunable light source that illuminates an extended region of the retina. Light reflected from the retina is corrected using a deformable mirror and wavefront sensor, and then interfered with light from a reference mirror. The x-y-λ spatio-spectral cube is converted by DFT into a 3D image of the retina, with amplitude and phase of light scattered from the corresponding tissue. Two volunteers were imaged. Various stimulus patterns were used, including multiple flashes separated by controllable delays, background illumination, and step-function illumination.
Results :
Cone OS phase exhibits dose-dependent contraction and elongation (1A), as previously reported. In some cases, responses to multiple flashes appeared to be linear combinations of single flash responses (1A, brown). These fit a mathematical model based on coupled overdamped oscillators (1A,B). Model parameters correlated with stimulus energy (1C-F). The model predicts some multiple-flash responses, confirming apparent linearity. To a series of 10 flashes given at 20 Hz, OS contractions diminish, as when flashes are presented against adapting backgrounds. Responses to step-function stimuli bore similarity to integrations of single-flash responses.
Conclusions :
A model was proposed that appears to fit the biphasic ORG response in cones. Parameters of this model exhibit dose-dependence and may offer distinct utility in quantifying the effects of experimental variables, including dark and light adaptation, circadian factors, or the presence of retinal disease.
Acknowledgements: NEI grants R01EY031098, R01EY026556, and R01EY033532, CNpQ grants.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.