Transition From Film to Digital Mosaic Grading: Longitudinal Studies of the Ocular Complications of AIDS (LSOCA)

Abstract

Purpose: : To compare evaluation of cytomegalovirus retinitis (CMVR) lesion from mosaiced digital images and film photographs, in order to test equivalency between image media and grading procedures in LSOCA.Background: Many clinics participating in the LSOCA clinical trial have transitioned from film to digital retinal photography during the study period. Consequently, the Fundus Photograph Reading Center (FPRC) has developed a new digital grading procedure for the on-going study.

Methods: : Film and digital photographs of 15 LSOCA patients were taken at the same visit from a sample of clinics. Topcon IMAGENET was used to store the original wide angle digital fundus images and the images were exported to Photoshop to construct a mosaic of the entire field of view. The mosaic was constructed by positioning 9 fields in anatomic location, using anatomic landmarks to facilitate alignment. Once the mosaic was completed, a digital version of the LSOCA grid was placed over the entire quasi-mosaic. The grading procedure included evaluation for the presence of CMVR lesions, assessment of the maximum activity of CMVR, and measurements of distance of the lesion to the optic disc and to center of the macula using Photoshop measuring tools. Area involved by CMVR was assessed by counting the number of blocks in the grid which were involved with CMVR.

Results: : Of the 15 film and digital images obtained, 5 did not show evidence of CMVR and hence were not mosaiced. 10 eyes with CMVR were graded independently at different sessions to minimize recall. The intraclass correlation coefficient (ICC) between film and digital images was 0.999 (95% confidence interval [CI], 0.994-1.0) for distance to disc, 0.998 (95% CI, 0.993-1.0) for distance to macula and 0.995 (95% CI, 0.982-0.999) for the lesion area. There was no evidence of systematic difference between imaging media for presence/absence and maximum activity.

Conclusions: : Our results show that mosaic grading for LSOCA is comparable to film grading ensuring continuity of grading data in the transition from film to digital imaging.

Keywords: AIDS/HIV • cytomegalovirus • imaging/image analysis: clinical