A Hybrid DE-RGSO-ELM for Brain Tumor Tissue Categorization in 3D Magnetic Resonance Images

Table 9

Comparison of various tumor classification methods.

SPL dataset	Traditional approach [32]		kNN approach [5]		(IELM approach) [28]		Proposed EM-RGSO
	Sensitivity	Specificity	Sensitivity	Specificity	Sensitivity	Specificity	Sensitivity	Specificity
Case 1	100	100	100	100	99.8	100	100	100
Case 2	82.2	99.9	94.7	99.9	97.8	99.8	98.20	99.9
Case 3	93.6	100	79.8	100	94.56	99.8	95.61	100
Case 4	73.3	100	73.3	100	72.4	100	73.6	100
Case 5	79.3	99.9	83.7	100	80.6	99.6	81.2	100
Case 6	90.5	99.9	96	99.9	93.5	99.6	94.9	99.9
Case 7	94.6	99.9	94.7	99.9	94.2	99.8	94.1	99.6
Case 8	73.6	99.9	93.1	99.9	88.6	99.4	88.1	99.9
Case 9	91.8	99.9	99.3	99.8	96.8	99.5	97.92	99.9
Case 10	97.8	100	95.6	99.8	96.8	99.6	97.4	99.8
Average	87.67	99.9	91.02	99.9	91.506	99.71	92.103	99.9