Diagnostic Validity of Macular Ganglion Cell-Inner Plexiform Layer Thickness Deviation Map Algorithm Using Cirrus HD-OCT in Preperimetric and Early Glaucoma

Abstract

To evaluate the diagnostic validity of macular ganglion cell-inner plexiform layer (mGCIPL) thickness deviation map algorithm using Cirrus high definition-optical coherence tomography to discriminate between normal controls and patients with preperimetric or early glaucoma. Seventy-two normal controls, 37 patients with preperimetric glaucoma and 70 patients with early glaucoma were enrolled. mGCIPL thickness and peripapillary retinal nerve fiber layer (pRNFL) thickness were measured by Cirrus high definition-optical coherence tomography. Areas showing abnormal color coding were obtained by customized Image J software calculating the number of abnormal superpixels at 1% and 5% level in each deviation map of measurements (GCIPL-DM1, GCIPL-DM5, RNFL-DM1, RNFL-DM5). The area under the receiver operating characteristic curve (AROC) of each parameter was calculated to provide diagnostic ability between normal controls and patients with preperimetric or early glaucoma. AROCs of the deviation map algorithms were higher than those of other parameters. Parameter with the best AROC was the GCIPL-DM5 (0.920 and 0.968) in both preperimetric and early glaucoma. The sensitivities of the GCIPL-DM5 at 80% and 95% specificities were 92% and 68% in preperimetric glaucoma and 98% and 90% in early glaucoma, respectively. Pairwise comparisons between AROCs of parameters from deviation map algorithms did not show statistically significant differences. mGCIPL thickness deviation map showed good diagnostic ability in detecting preperimetric and early glaucoma, and it was comparable with pRNFL thickness deviation map. Our findings suggest that it can be an important parameter in detecting subtle glaucomatous structural change.