Table 2.

Mean metrics for comparing our method to expert annotators.

Mean metrics	Automated VS expert annotation	Agreement of three experts (interobserver)
IoU - Jaccard index $\Uparrow$ ($\mathrm{\%}$)	80.02${}^{\mathrm{a}}$	85.22${}^{\mathrm{b}}$
Dice score $\Uparrow$ ($\mathrm{\%}$)	86.40${}^{\mathrm{a}}$	90.81${}^{\mathrm{b}}$
Hausdorff distance $\Downarrow$ (mm)	4.68${}^{\mathrm{a}}$	2.65${}^{\mathrm{b}}$
Mean surface distance $\Downarrow$ (mm)	2.10${}^{\mathrm{a}}$	1.12${}^{\mathrm{b}}$

For each metric, differences are statistically significant ($p<0.001$, also denoted with different lettering in superscripts). $\Uparrow$ indicates higher-the-better metrics, while metrics $\Downarrow$ are lower-the-better.