Can supercell analysis diagnose Birdshot uveitis?

Research details

  • Type of funding: Fight for Sight / Birdshot Uveitis Society Small Grant Award
  • Grant Holder: Dr Graham Wallace
  • Institute: University of Birmingham
  • Region: West Midlands
  • Start date: March 2015
  • End Date: March 2016
  • Priority: Early detection
  • Eye Category: Ocular inflammatory


Birdshot uveitis is a rare form of uveitis that affects around 200-300 people in the UK. It’s hard to diagnose because the symptoms are subtle early on and ophthalmologists aren’t familiar with it because it’s so rare. There is no diagnostic test, yet.

In this project the research team is trying a machine learning method known as ‘super cell analysis’ to see if they can help speed up Birdshot diagnosis. The computer program takes in data about hundreds of blood cells taken from patients and from healthy ‘controls’ without any eye disorder. It looks at lots of different features of the cells and tries to separate out the information that comes from patients. The method has already been used successfully by researchers to tell the difference between people with the eye condition Behçet syndrome and people with another condition called ocular sarcoidosis.

At the end of the project we should have a better idea of how well this method can tell the difference between Birdshot and healthy controls or people with other eye conditions. We should also know more about how the test results relate to the signs and symptoms people show in the clinic. Ultimately this could lead to a reliable diagnostic test that would mean people could start treatment earlier.

Find out more about Birdshot Uveitis Society