Automated, unbiased optoretinography reveals comprehensive nanoscopic dynamics of the outer retina in rodents

Small animals, such as rodents, serve as an attractive option for investigating the intrinsic process of photoreceptor degeneration because of their wide availability and versatility in disease models and gene manipulations. However, there was a lack of an objective and quantitative approach to measure their outer retina dynamics while preserving spatial heterogeneity. Here, we demonstrate an automated, unbiased approach for functional retinal imaging in rodents based on unsupervised machine learning. Our method automatically searches for and classifies nanoscopic cellular dynamics from obscure speckle patterns captured by a low-cost, phase-sensitive optical coherence tomography. Using this approach, we revealed highly reproducible Type-I and Type-II signals in rodents related to different parts of the outer retina. The fast Type-I signal was not reported previously and we hypothesized that it originated from the movement of rod outer segments. We also characterized the light-induced response of the outer retina under scotopic and photopic conditions, and demonstrated en-face mapping of the outer retina function in an extended field of view (12°), analogous to multifocal electroretinograms, but only with a single shot and yielding much higher spatial resolution. Our approach can be widely applied to investigating tissue-specific retinal dynamics across animal models, as well as facilitating clinical translations for the early detection of neurodegenerative diseases. ### Competing Interest Statement The authors have declared no competing interest.