Semi-automated brain tumor segmentation on multi-parametric MRI using regularized non-negative matrix factorization

Table 2 Mean Dice-scores [%] for NMF without regularization, with spatial regularization, with spatial and sparse regularization but without morphological post-processing, with spatial and sparse regularization on the cMRI data only, and with spatial and sparse regularization on the full MP-MRI dataset

	Active tumor [%]			Tumor core [%]			Whole tumor [%]
	25_th prcntile	Mean	75_th prcntile	25_th prcntile	Mean	75_th prcntile	25_th prcntile	Mean	75_th prcntile
\(\mathbf {NMF_{\text {no\_reg}}}\)	58	63	72	66	72	79	76	80	84
N M F _spatial	55	60	70	65	71	79	76	80	85
\(\mathbf {NMF_{\text {no\_postproc}}}\)	54	59	67	62	68	73	71	73	77
N M F _cMRI	60	65	74	67	72	79	73	77	82
\(\mathbf {NMF_{\text {spatial\_sparse}}}\)	60	65	74	68	74	80	77	80	85

For each tissue class, the first column reports the mean of the 25^th percentile Dice-score across all patients, the second column the mean of the mean Dice-score across all patients and the third column the mean of the 75^th percentile Dice-score

ISSN: 1471-2342