Active

March 2025 - February 2026

Applying AI to screening for Birdshot Chorioretinopathy and predicting response to treatment

Research Details

  • Type of funding: Fight for Sight / Birdshot Uveitis Society Small Grant Award
  • Grant Holder: Dr Nikolas Pontikos
  • Region: London
  • Institute: UCL Institute of Ophthalmology
  • Priority: Early diagnosis
  • Eye Category: Ocular inflammatory

Brief Lay Background

Birdshot chorioretinopathy (BCR) is an autoimmune condition affecting the eye that can lead to visual loss and blindness. We have a limited understanding of how genes affect the course of BCR, how they influence the disease and whether the appearance of the retina or response to treatments can be affected.

What problem/knowledge gap does it help address?

Many patients with BCR experience long delays in reaching a diagnosis, partly due to the rarity of the condition, but also because it can be subtle in appearance and easily missed or confused with other diseases. We also cannot currently predict which patients with BCR will be likely to respond to standard treatments that suppress the immune system.

We will use data from a new clinical study called ‘BIRD-SET’ to develop approaches to screen and detect BCR from retina photographs and identify patterns in genes and medical records that may help to better manage patient care. The BIRD-SET study will have data from 180 patients that have undergone a type of genetic testing called ‘whole exome sequencing’.

Aim of the research project

We aim to develop an artificial intelligence screening tool to identify BCR from colour photographs of the eye, before looking for features from medical records, images and genetic testing that could predict the responsiveness to common immune system treatments. 

Potential impact on people with sight loss

The project should lead to a new tool that can be used in hospitals and ultimately community opticians, to detect BCR earlier and not miss the diagnosis or confuse it with another condition. Based on analysing the genetics of BCR patients alongside retinal images and medical details, new ways to predict how patients will respond to early treatments could be achieved.

Latest from our research blog