Learning rich features with hybrid loss for brain tumor segmentation

Table 4 Performance comparison of different ensemble models on the test set

Method	DSC			PPV			Sensitivity
Method	Complete	Core	Enh	Complete	Core	Enh	Complete	Core	Enh
DeepMedic (ensemble) [12]	0.845	0.667	0.633	0.833	0.861	0.632	0.889	0.599	0.673
DeepMedic + CRF (ensemble) [12]	0.849	0.667	0.634	0.853	0.861	0.634	0.877	0.600	0.674
FCNN + 3D CRF (fusing) [13]	0.84	0.73	0.62	0.89	0.76	0.63	0.82	0.76	0.67
Proposed + \(\mathrm{H}{L}_{cb\_\_rl\_loss}\)	0.86	0.73	0.61	0.86	0.76	0.60	0.88	0.76	0.65
Proposed + \(\mathrm{H}{L}_{ce\_\_rl\_loss}\)	0.86	0.72	0.60	0.86	0.76	0.56	0.88	0.74	0.70

ISSN: 1472-6947