18.11.2026

Deep learning image processing for GelCount™

Introducing a powerful image processing module for the detection of luminal organoids

Image taken by GelCount 2 of a classic dome of luminal organoids in a 6-well plate, processed by our new deep learning module to reveal a total of 265 organoids varying in diameter from 75 µm to 850 µm.


Responding to the recent explosion in the use of
in vitro organoid models in human-relevant multicellular drug discovery, regenerative medicine, and disease modelling research we have developed an image processing module for GelCount, specifically to address the requirements of organoid detection.

The traditional algorithmic image processing approach employed by GelCount was originally developed for the detection of spheroids and colonies in the classic colony formation assay, however it is not optimized for the reliable detection of luminal organoids. 

These are commonly cultured in 3-dimensional agar ‘dome’ structures and present as ‘hollow’, highly overlapping and heterogeneously sized objects. 

We have employed a deep learning approach using a dataset of luminal organoids imaged by GelCount to dramatically improve the detection of these objects.

Commenting on the development, Andy Obeid, CEO and founder of Oxford Optronix commented, “We’re excited to leverage modern day AI-based tools to offer this exciting new capability for our GelCount imaging platform. We see this as a new beginning for GelCount”.

The new capability is available immediately as a paid ‘bolt-on’ module compatible with our version 2 image processing software for users of our latest-generation GelCount™ 2 device. 

GelCount can accommodate up to 4 multi-well plates at a time, yields object count and object diameter statistics directly into Excel® , and is compatible with a robotic loading solution recently developed in collaboration with Astech Projects Ltd., UK 

 

Bookmark our news feed or follow us on X for the latest news and updates from Oxford Optronix.

Read More

Introducing OxyData™ v2 acquisition software

Affordable in-house data logging for OxyLite™ and OxyFlo™

Image7

Robotic sample loading for GelCount™

For high throughput applications

Image1 GC logo crop

Standout 2026 articles citing our monitors

Serving the scientific community

Generic scientifc article image

Your choice regarding cookies on this site. For more information regarding the cookies we use, please see our Privacy Policy