Table 2.

Comparison of evaluation metrics of various typical models.

Models	DSC ↑	HD95 (mm) ↓	Precision ↑	Recall ↑
3D UNet	0.7172 ± 0.0955	5.7052 ± 26.80	0.7354 ± 0.1185	0.7223 ± 0.1317
Isensee et al. [24]	0.7693 ± 0.0893	2.4136 ± 3.5872	0.7633 ± 0.1143	0.7924 ± 0.1163
2D Isensee et al. [24]	0.6520 ± 0.1463	7.9084 ± 10.2189	0.8081 ± 0.1134	0.5763 ± 0.1846
Lin et al. [23]	0.7754 ± 0.0954	2.6832 ± 4.7794	0.7695 ± 0.1177	0.7983 ± 0.1197
TransBTS	0.7744 ± 0.0929	8.6436 ± 87.3025	0.7583 ± 0.1107	0.8080 ± 0.1221
UNETR	0.7683 ± 0.0944	3.0822 ± 6.2588	0.7687 ± 0.1184	0.7850 ± 0.1199
VT-UNet	0.7758 ± 0.0906	2.5433 ± 5.1414	0.7690 ± 0.1073	0.7989 ± 0.1203
MPU-Net	0.7693 ± 0.0963	3.1479 ± 7.7695	0.7817 ± 0.1143	0.7747 ± 0.1279
LVPA-UNet	0.7907 ± 0.0937	1.8702 ± 2.8491	0.7929 ± 0.1088	0.8025 ± 0.1201