Table 3. The mean unsigned border position errors (mean ± SD μm) with respect to three raters.

Surface	Auto vs. Rater 1	Auto vs. Rater 2	Auto vs. Rater 3	Auto vs. Rater 1	Auto vs. Rater 2	Auto vs. Rater 3
ILM	1.41 ± 0.41	1.43 ± 0.40	1.36 ± 0.32	1.40 ± 0.31	1.39 ± 0.34	1.38 ± 0.42
RNFGC-IPL	3.41 ± 1.70	4.61 ± 2.49	3.41 ± 1.87	3.13 ± 1.10	3.38 ± 1.40	2.88 ± 0.77
IPL-INL	2.49 ± 0.94	3.72 ± 1.60	2.49 ± 0.65	2.45 ± 0.68	3.22 ± 0.99	2.16 ± 0.48
INL-OPL	1.72 ± 0.59	2.13 ± 0.66	2.21 ± 0.75	2.50 ± 0.54	2.70 ± 0.64	2.09 ± 0.61
OPL-ONL	2.06 ± 0.59	2.64 ± 1.29	2.32 ± 0.44	2.45 ± 1.27	3.12 ± 1.35	2.19 ± 1.18
ELM	1.66 ± 0.98	2.06 ± 1.08	1.57 ± 1.00	–	–	–
EZ-Top	1.75 ± 1.11	2.49 ± 1.41	2.59 ± 1.18	–	–	–
EZ-Bottom	1.30 ± 1.24	3.03 ± 1.58	1.46 ± 1.39	–	–	–
OS-RPE	2.10 ± 1.25	2.31 ± 1.46	2.16 ± 1.34	–	–	–
BM	1.71 ± 1.46	1.84 ± 1.51	1.83 ± 1.46	2.52 ± 1.16	2.47 ± 1.14	2.43 ± 1.02
Average	1.96 ± 0.62	2.63 ± 0.94	2.14 ± 0.62	2.41 ± 0.56	2.71 ± 0.73	2.19 ± 0.49