Fig. 3: Under-estimation of severity by a single visual field test due to missed central (a) or peripheral (b) involvement.

a An example of the gradings based on the 24-2 under-estimating severity due to missed central involvement. Based only on the 24-2 (i. red rectangle) the graders classified severity as moderate due to superior hemifield involvement (black arrow). However, according to the RS (iv and v—black rectangle), this eye had superior hemifield involvement on the RNFL probability map (iv) based on 24 aS-aF locations and 4 aS-aF-aF clusters (blue outline), and central involvement on the GCL probability map (v) based on 4 aS-aF locations (yellow arrows) and 3 aS-aF-aF clusters (yellow outline) within the central 5 degrees. Therefore, the RS severity was classified as advanced. Note that when the graders evaluated the severity for this eye based on the 10-2 (iii) alone as well as based on the combination of OCT (ii) with either 24-2 or 10-2, they agreed with the RS and classified the severity as advanced due to central involvement, in agreement with the damage on the GCL probability map (red arrows). b An example of the gradings based on the 10-2 under-estimating severity due to missed hemifield involvement. Based only on the 10-2 (iii, red rectangle) the graders classified severity as mild. However, according to the RS, this eye had superior hemifield involvement on the RNFL probability map (iv) based on 2 aS-aF locations from the 24-2 pattern deviation map, in agreement with the superior arcuate defect on the RNFL probability map (blue arrows). Therefore, the RS severity was classified as moderate. Note that when the graders evaluated the severity for this eye based on the 24-2 (i) alone as well as based on the combination of OCT (ii) with either 24-2 or 10-2, they agreed with the RS and classified the severity as moderate due to superior hemifield involvement.