Emulated retinal image capture (ERICA) to test, train and validate processing of retinal images

Laura K Young,Hannah E Smithson

doi:10.1038/s41598-021-90389-y

Laura K Young, Hannah E Smithson

Open Access

https://doi.org/10.1038/s41598-021-90389-y

Copy DOI

Journal: Scientific Reports	Publication Date: May 27, 2021
Citations: 3	License type: open-access

Affiliation: Newcastle University, University of Oxford

Abstract

High resolution retinal imaging systems, such as adaptive optics scanning laser ophthalmoscopes (AOSLO), are increasingly being used for clinical research and fundamental studies in neuroscience. These systems offer unprecedented spatial and temporal resolution of retinal structures in vivo. However, a major challenge is the development of robust and automated methods for processing and analysing these images. We present ERICA (Emulated Retinal Image CApture), a simulation tool that generates realistic synthetic images of the human cone mosaic, mimicking images that would be captured by an AOSLO, with specified image quality and with corresponding ground-truth data. The simulation includes a self-organising mosaic of photoreceptors, the eye movements an observer might make during image capture, and data capture through a real system incorporating diffraction, residual optical aberrations and noise. The retinal photoreceptor mosaics generated by ERICA have a similar packing geometry to human retina, as determined by expert labelling of AOSLO images of real eyes. In the current implementation ERICA outputs convincingly realistic en face images of the cone photoreceptor mosaic but extensions to other imaging modalities and structures are also discussed. These images and associated ground-truth data can be used to develop, test and validate image processing and analysis algorithms or to train and validate machine learning approaches. The use of synthetic images has the advantage that neither access to an imaging system, nor to human participants is necessary for development.

Highlights

AO retinal imaging provides unprecedented opportunities for fundamental vision r esearch[5,7]
At present the most commonly used AO retinal imaging system is the adaptive optics scanning laser ophthalmoscope (AOSLO)[5], with AO flood illumination used by many groups, and AO optical coherence tomography systems being rapidly developed and used
We present ERICA (Emulated Retinal Image CApture), the first end-to-end simulation of AOSLO data capture generating a realistic output, with specified ground-truth data, comparable to the stream of images obtained from human participants

Summary

Introduction

AO retinal imaging provides unprecedented opportunities for fundamental vision r esearch[5,7]. Pioneered by Roorda et al in 2 00210, the AOSLO generates images by raster-scanning a point source over an area of the retina and measuring the reflected light after focusing it through a confocal pinhole. It is this type of system that we simulate in the current work. An example relevant to the current work is the exploitation of the motion distortions introduced by scanning during retinal imaging (with and without AO correction) This allows measurement of the microscopic eye movements made during fixation to understand the role that they play in spatial v ision[15,16,17,18,19,20]. The robust collection of good quality data from AO retinal imaging systems, as well as robust data processing automation, is clearly important for these applications

Methods

Results

Conclusion