Table 3.

Histology data. Dice (in %) achieved at five different labelling budgets (in % of total boundary length).

Bound. length	35%	40%	50%	70%	90%
$\mathit{\text{BD}}\mathit{\alpha }$	91.67 ± 6.57	94.44 ± 5.29	98.04 ± 2.22	99.79 ± 0.22	99.95 ± 0.04
$\mathit{\text{BD}}\mathit{\alpha }\_\mathit{\text{CE}}$	87.99 ± 8.59	92.17 ± 6.31	96.69 ± 3.68	99.64 ± 0.39†	99.96 ± 0.02†
$\mathit{D}\mathit{\alpha }$	79.32 ± 13.00	81.90 ± 12.96	91.94 ± 10.12	99.14 ± 2.59	99.96 ± 0.03
$\mathit{\text{HBD}}\mathit{\alpha }$	89.04 ± 8.26†	92.21 ± 7.17	96.04 ± 3.96	99.38 ± 0.92†	99.92 ± 0.10†
$\mathit{\text{HB}}$	85.91 ± 8.23	89.49 ± 7.63	92.68 ± 5.50	97.59 ± 2.13	99.23 ± 0.73
$\mathit{\text{HD}}\mathit{\alpha }$	71.28 ± 9.44	72.29 ± 9.18	75.57 ± 8.11	86.77 ± 6.47	97.77 ± 2.14
$\mathit{H}$	71.10 ± 9.48	71.55 ± 9.28	72.34 ± 9.09	73.92 ± 8.68	80.80 ± 6.85
$\mathit{H}\_\mathit{\text{dropout}}$	70.25 ± 9.95	70.82 ± 9.65	71.27 ± 9.42	72.91 ± 9.05	80.36 ± 7.15
$\mathit{\text{Rand}}$	72.76 ± 9.02	73.70 ± 8.74	75.25 ± 8.38	79.40 ± 7.30	86.85 ± 5.46
$\mathit{\text{Uniform}}$	51.16 ± 13.64	55.24 ± 13.15	62.46 ± 13.84	75.80 ± 14.64	93.53 ± 5.49

The dagger $†$ indicates that $p>0.001$ for a Wilcoxon signed-rank test test comparing the method at hand against ours.