Multimodal MRI Brain Tumor Image Segmentation Using Sparse Subspace Clustering Algorithm

Table 2

Comparison of multimodal image segmentation results.

Experimental sample		Index
		Dice	Jaccard	Precision	Recall
Malignant tumor	1	0.8923	0.8042	0.9627	0.8287
	2	0.8596	0.7752	0.9748	0.7789
	3	0.8385	0.7431	0.9263	0.7821
	4	0.9507	0.8785	0.9874	0.9264
	5	0.9310	0.7886	0.9845	0.7954
	6	0.8746	0.7964	0.9678	0.8103
	7	0.9152	0.8522	0.9034	0.9371
	8	0.8386	0.7371	0.9976	0.7352
	9	0.8731	0.7796	0.9948	0.7911
	10	0.8694	0.7649	0.9563	0.8002
	11	0.8627	0.7628	0.9915	0.7832
	12	0.7016	0.5364	0.9997	0.5349
	13	0.8018	0.6742	0.9306	0.7132
	14	0.8220	0.6976	0.9736	0.7120
	15	0.8129	0.6842	0.9637	0.7058
Bright tumor	1	0.7961	0.6425	0.9264	0.6779
	2	0.8129	0.8413	0.9779	0.8646
	3	0.9298	0.6624	0.9836	0.6732
	4	0.9401	0.6830	0.7375	0.9118
	5	0.9228	0.8698	0.9990	0.8769
	6	0.9418	0.8891	0.9862	0.8996
	7	0.7147	0.5510	0.9996	0.5534
	8	0.7753	0.6256	0.9676	0.6394
	9	0.9027	0.8244	0.9834	0.8426
	10	0.8624	0.7632	0.9623	0.7824
Mean		0.8577	0.7451	0.9615	0.7743