Table 1.

Intraobserver Annotation Variability Analysis

Specialist	F1-Box Score	F1-Mask Score	MAE
A	0.070	0.831	1.714
B	0.167	0.884	1.286
C	0.062	0.692	1.667
D	0.127	0.694	1.095
E	0.055	0.846	3.286
Mean	0.096	0.789	1.810

Abbreviation: MAE, mean absolute error.

For each specialist, we calculated the F₁ scores (F₁-box and F₁-mask) and absolute error in every image pair and then averaged the results.