Abstract
Purpose :
We developed a hyperemia analysis software which is able to evaluate the levels of hyperemia. In this study, we aimed to evaluate the severity of conjunctival hyperemia by clinical grading and the hyperemia analysis software, and to investigate the correlation of data between these two methods.
Methods :
Samples used were conjunctival photos from 13 patients exhibiting different severity of hyperemia. Conjunctival photos were taken by a slit lamp (model SL-D7; Topcon, Tokyo, Japan). Resolution of photos was 2000 * 1312 pixels. The evaluation was focused on the restricted area (region of interest: ROI) in the temporal side of bulbar conjunctiva (800 * 800 pixels). In the case of hyperemia analysis software, the percentage pixel coverage of blood vessels in the ROI was evaluated. Clinical grading was determined following the Japanese guideline for allergic conjunctival diseases. The severity was classified into either no significant (-), mild (+), moderate (++) or severe (+++). Evaluation was conducted by 9 board members of Japanese Ocular Allergy Society. The area for evaluation by clinical grading was just in accord with that by hyperemia analysis software. Spearman rank-correlation coefficient was used for statistical analysis (R 3.2.2).
Results :
Clinical grading data by observers were assumed to be arithmetic average. Arithmetic averages of clinical grading data were correlated with the pixel coverage (R=0.953; 95% Cl 0.85-0.99 P<0.001).
Conclusions :
We suggest that pixel coverage of data obtained by hyperemia analysis software reflects clinical grading data of conjunctival hyperemia.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.