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

Table 5

Clinical and imaging characteristics of 70 patients using Student’s 𝑡-test for 3D VOI.

Features	AUC value	𝑃 values (Student’s 𝑡-test)	Ranking
Energy	0.91	4.28𝐸^−4	3
Entropy	0.87	3.55𝐸^−6	5
Contrast	0.83	3.02𝐸^−2	10
Homogeneity	0.56	8.85𝐸^−6	27
Correlation	0.63	2.74𝐸^−2	23
Variance	0.90	1.20𝐸^−5	4
Inverse difference moments	0.61	8.46𝐸^−6	24
Sum mean	0.84	5.56𝐸^−3	9
Sum variance	0.86	3.13𝐸^−5	6
Sum entropy	0.72	3.66𝐸^−3	16
Difference entropy	0.85	4.76𝐸^−8	8
Infm. correlation 1	0.64	5.05𝐸^−5	22
Infm. correlation 2	0.71	4.67𝐸^−5	17
Maximum correlation coefficient.	0.65	3.75𝐸^−2	21
Short run emphasis	0.85	7.92𝐸^−8	7
Long run emphasis	0.69	1.562𝐸^−3	19
Low gray-level run emphasis	0.68	4.56𝐸^−2	20
High gray-level run emphasis	0.81	3.21𝐸^−2	11
Short run low gray-level run emphasis	0.77	5.12𝐸^−1	14
Short run high gray-level run emphasis	0.79	6.10𝐸^−3	12
Long run low gray-level run emphasis	0.61	3.88𝐸^−8	25
Long run high gray-level run emphasis	0.72	1.459𝐸^−4	15
Gray-level nonuniformity	0.93	3.15𝐸^−2	2
Run-level nonuniformity	0.59	1.49𝐸^−5	26
Run percentage	0.70	1.47𝐸^−8	18
Gradient vector parameter	0.93	6.2𝐸^−2	1
Gradient vector orientation	0.78	4.2𝐸^−3	13