Automated leakage analysis for monitoring posterior uveitis

This review discusses the use of automated analysis of retinal vascular leakage to monitor patients with posterior uveitis.

Background

Retinal vascular leakage is an important clinical marker of uveitis involving the posterior segment. Control of peripheral vascular leakage (PVL) is important for good outcomes in these patients and could therefore be a useful visual and quantitative endpoint. However, it can be difficult to assess leakage in a clinical setting.

Research suggests ultra-widefield (UWF) angiography may be a more effective method to monitor PVL than standard fluorescein angiography. Software for automated analysis of vascular leakage on UWF angiography could be useful for managing patients with inflammatory ocular disease. Several teams are actively working on developing quantitative leakage analysis algorithms.

Findings

Dr. Ehlers and colleagues published an algorithm to perform automated quantification of leakage in diabetic patients and developed a leakage index. The software can distinguish true leakage from background fluorescence based on a minimum intensity that is adjacent to vasculature. The protocol demonstrates a strong correlation between manual and automated quantification of vascular leakage.

Drs. Venkat and Sharma used this algorithm to analyze a patient with uveitis treated with adalimumab. The pre- and post-treatment UWF images demonstrated a clear reduction in leakage after 6 months of treatment. The automated software quantified a 13.6% reduction in leakage during the treatment period.

Clinical significance

Automated leakage analysis on UWF angiography would make it easy to monitor and measure vascular leakage as a clinical endpoint in patients with uveitis.