Cone photoreceptor analysis got faster, more objective, and scalable

Characterizing the cone mosaic from adaptive optics images has long been a bottleneck. Time-consuming, and difficult to standardize. AOBOX™ is designed to change that.

Erasmus MC and Imagine Eyes introduce AOBOX

AOBOX is an AI-based application for analyzing ultrahigh-resolution images from adaptive optics retinal cameras, co-developed by Erasmus MC and Imagine Eyes. It is now in active deployment at 15 leading ophthalmology centers worldwide. This stand-alone software enables reproducible, objective characterization of the cone photoreceptor mosaic – at a scale and speed that less automated workflows simply cannot match.
It is intended for investigational use.

What AOBOX delivers

  • Automated cone photoreceptor detection, with optional human annotation
  • Heat maps of cone metrics, including cone density and inter-cell spacing
  • Average density/spacing across customized regions of interest – providing highly sensitive biomarkers of retinal cellular integrity
  • Batch processing of large datasets, including multi-image montages – built for high-throughput research workflows
  • Full compatibility with the RTX1™ AO retinal camera
  • Local installation, so sensitive patient data stay securely within your institution

Built to grow with the field

AOBOX is more than a tool – it is part of an open innovation initiative shared across the RTX1 user community. Early users are already gathering annotated datasets that will help refine the AI and expand its biomarker capabilities over time. As a therapy developer or clinical researcher, you are in a position to shape what this technology becomes. We welcome both contributors and users.

Meet us at ARVO 2026

Discover AOBOX in person at booth #7002. We would be glad to walk you through the software, discuss your research needs, and explore how AOBOX can integrate into your workflow.

Schedule a meeting at ARVO

 

AOBOX development was supported by Health Holland, Top Sector Life Sciences & Health under grant agreement no. EMCLSH22014 (AO-VISION). Reference publication on the AI model for cone detection: Wooning et al., Ophthalmology Science, 2024.

AOBOX application is intended for investigational use only.

 

Explore rtx1 capabilities About AO-Vision project