New diagnostic offers hope to sufferers of serious eye infections that pose biggest risk to contact lens wearers
Some of the most devastating eye infections leading to possible blindness and to which contact lens wearers are particularly vulnerable have until now proved difficult to diagnose quickly enough for effective treatment.
Now researchers at the University of Bradford and ophthalmologists at Manchester Royal Eye Hospital have developed an automatic diagnostic system to improve the detection, and treatment, of two of the most serious eye infections.
Infectious keratitis (Acanthamoeba keratitis and fungal keratitis or Fusarium) attacks the cornea and can lead to irreversible complications and even blindness.
A significant risk factor is contact lens wear for acanthoamoeba infections, especially in unhygienic circumstances such as swimming while wearing contact lenses, the use of non-sterile lens solutions, and insufficient disinfection practices. In the United Kingdom, Europe, Hong Kong, and the USA, the estimated infection rate is 1.2 per million adults per year, but this rises to between 0.2 and 1 per 10,000 amongst contact lens wearers.
Successful treatment needs early diagnosis and intervention but accurate and rapid identification of the infectious agent involved and proper management of corneal ulcers are challenging clinical problems. Microorganisms need time to be detected using corneal cultures and smears, which are currently considered to be the gold standard diagnostic tools for infectious keratitis, and they may not be detected because of inadequate sample material, delay in performing the investigations, deeply seated lesions, and previous use of empirical antimicrobial treatments.
The affected areas of Acanthamoeba keratitis (cysts) and Fusarium (fungal filaments) need to be detected first using confocal microscopy (an optical imaging technique for increasing optical resolution and contrast) and then need to be tracked in the sequence of images that are captured from different depths of the patient's cornea. However, manual detection and tracking of those signs in confocal images are subjective, time-consuming and error-prone.