A New Scale (Penn Scale) for Measuring Bulbar Conjunctival Redness from Digital Photographs

Abstract

Purpose : Current scales for measuring bulbar conjunctival redness have limitations when used with digital photographs. This study sought to develop a scale suited for digital images (Penn Scale) and compare the grading results from it to those using the VBR (Validated Bulbar Redness) scale.

Methods : A digital image database of 4889 color corrected nasal and temporal images of eyes with varying degrees of drug-induced bulbar conjunctival redness was reviewed to identify the least (score 1) and most severe redness (score 10), as well as 8 intermediate levels of redness. These 20 digital images, ten each of nasal and temporal, comprise the Penn scale, the linearity of which was assessed with an established image processing algorithm. 100 randomly selected images from the database were assessed by 3 professional non-physician graders using the Penn scale and the VBR scale. Agreement among graders was assessed with Kappa statistics using Cicchetti-Allison weights. The ease of use of the two scales was also documented.

Results : The Penn scale showed an almost perfect linearity of redness (Figure 1). The weighted Kappas for agreement between any two graders ranged from 0.57 to 0.73 for the Penn scale and from 0.38 to 0.66 for the VBR scale (Table 1). Among all three graders the weighted Kappa was 0.39 for the Penn scale and 0.14 for the VBR scale. In contrast to the VBR scale, the Penn scale allowed direct comparison of digital to digital images, could be used in a dimly lit environment conducive for evaluating digital images, had scaled bulbar conjunctival images for both temporal and nasal sides, and permitted reference scale images to be magnified to the same size as the images to be graded.

Conclusions : The Penn scale demonstrated several advantages over the VBR scale for the subjective measurement of bulbar redness from digital images.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.