A state-of-the-art technique to perform cloud-based semantic segmentation using deep learning 3D U-Net architecture

Table 7 List of methods with a high accuracy and dice score

Dataset	Segmentation method	Dice score	Publication year	References
BRATS 2013 BRATS 2015	FCNNs and CRFs	0.83 0.82	2017	[8]
BRATS 2013	CNN	0.80	2017	[9]
BRATS 2015	DeepMedic + CRF	0.71	2017	[10]
BRATS 2019	CNN	0.84	2020	[11]
BRATS 2015	FCN	0.89	2019	[12]
BRATS 2013	CNN	HG 0.943 LG 0.950	2019	[13]
BRATS 2015	RDM-Net	0.73	2019	[14]
BRATS 2015	CNN + TVS	0.857	2018	[15]
BRATS 2013	Hybrid CNN	0.86	2019	[16]
BRATS 2015	WRN-PPNet	0.94	2019	[17]
BRATS 2019	Two-stage Unet	0.84	2020	[18]
BRATS 2019	DNN	0.85	2020	[19]
BRATS 2017 BRATS 2018 Oslo Dataset	3D U-Net	0.82 0.84 0.80	2020	[20]
BRATS 2020	Ensemble + post-processing	0.87	2021	[21]
BRATS 2018	C-ConvNet	0.90	2021	[1]
BRATS 2020	3D U-NET	0.95	2022	This Study

ISSN: 1471-2105