Abstract
Purpose :
To compare conventional metrics to a topographical method for combining optical coherence tomography (OCT) and visual field (VF) information in order to confirm early glaucoma.
Methods :
Widefield OCT scans and 24-2 and 10-2 VFs were obtained on 101 eyes of 101 patients with, or suspected, early open-angle glaucoma; eyes had 24-2 mean deviation better than -6dB. Two glaucoma specialists independently judged 44 eyes as “healthy” (H eyes) and 57 as “glaucomatous” (G eyes), based upon disc photos, 10-2 and 24-2 VF, OCT, max IOP, family history, refractive error and corneal thickness.[1] The best conventional VF metric for classifying eyes as abnormal was an abnormal 24-2 (GHT or PSD) and/or an abnormal 10-2 (PSD); the best conventional OCT metric was circumpapillary layer RNFL quadrant thickness.[2] For the metric (M) method, eyes were considered abnormal if one of these VF metrics and the OCT were abnormal. For the topographical (T) method, eyes were considered abnormal if 3 or more 10-2 or 24-2 points at ≤5%, or 2 at 1%, fell within regions of abnormality (red, ≤1%) on RGC+ and/or RNFL probability maps. In Fig. 1 (left panels), the regions abnormal on the 10-2 (within black borders) and 24-2 (red borders) are shown on OCT probability maps (right panels) with the symbols indicating the abnormal 10-2 (squares) and 24-2 (circles) locations falling within abnormal OCT abnormal (red) regions.
Results :
For the M method, there were 15 (26.3%) misses (i.e., 15 of the 57 G eyes were classified as normal). For the T method, there were only 3 (5.3%) misses, and 2 of these showed qualitative agreement (e.g., Fig. 2A). The methods falsely classified 1 (T) and M (0) of the 44 H eyes as abnormal. Fig. 2B shows results for an H eye with 24-2 PSD/GHT, 10-2 PSD and OCT (quadrant) all within normal limits, but with abnormal 10-2 VF locations topographically agreeing with RGC+ regions (open squares). Note: the 24-2 (10-2) metric by itself missed 13 (13) of the 57 G eyes; the associated T methods missed 13 (7).
Conclusions :
There is excellent agreement between VF and OCT measures of early glaucoma as long as a topographical method, not conventional metrics, is employed and both 24-2 and 10-2 VF abnormalities are considered. 1. Hood et al., TVST (2016); 2. Muhamad et al., J Glau (2017).
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.