A recent study found that AI models were nearly as accurate as eye specialists in identifying infections. (Photo: Getty Images)Weiquan Lin

AI can diagnose corneal infections just like an ophthalmologist

AI technology is on the verge of transforming eye care by aiding in the diagnosis of infectious keratitis, a major cause of corneal blindness, with accuracy comparable to eye specialists.

by · India Today

In Short

  • AI can help diagnose infectious keratitis, which causes blindness
  • AI models can differentiate between healthy eyes and eyes with infections
  • While AI was 93.2% specific in diagnosis, ophthalmologists were 89.6%

Artificial Intelligence (AI) may soon play a crucial role in diagnosing infectious keratitis (IK), a leading cause of corneal blindness worldwide.

Corneal blindness is an umbrella term for a group of eye conditions caused by scarring of the cornea, leading to vision loss or blindness.

A recent study found that AI models, specifically deep learning (DL) models, were nearly as accurate as eye specialists in identifying infections.

The study, led by Dr Darren Ting from the University of Birmingham, reviewed 35 studies that used DL models to diagnose infectious keratitis.

Corneal blindness is an umbrella term for a group of eye conditions caused by scarring of the cornea, leading to vision loss or blindness. (Photo: Getty Images)

Published in eClinicalMedicine, the research showed that AI had a diagnostic sensitivity of 89.2% and specificity of 93.2%.

In comparison, ophthalmologists demonstrated a sensitivity of 82.2% and specificity of 89.6%. The AI models analysed more than 1,36,000 corneal images, proving their potential to assist in medical settings.

"Our study shows that AI has the potential to provide fast, reliable diagnoses, which could revolutionize how we manage corneal infections globally, especially in regions with limited access to eye care specialists. This could reduce cases of preventable blindness," Dr Ting explained.

Research showed that AI had a diagnostic sensitivity of 89.2% and specificity of 93.2%. (Photo: Getty Images)

The AI models not only accurately identified infections but also distinguished between healthy eyes and various types of IK, such as bacterial and fungal infections.

While the results are promising, the study's authors highlighted the need for more diverse data and further testing to ensure reliability in clinical practice.

Infectious keratitis affects millions, especially in low- and middle-income countries with limited access to specialized care.

As AI continues to evolve, it could become a powerful tool in preventing corneal blindness worldwide, the authors of the study said.