Feature extraction from MRI ADC images for brain tumor classification using machine learning techniques

Table 7 Tumor types and percentages belonging to each benign and malignant categories

Category	WHO grading	Tumor type	Image slices	Percentage (%)
Benign	WHO I	Meningioma	262	43.38
		Schwannoma	135	22.35
		Pilocytic astrocytoma	13	2.15
		Hemangioblastoma	16	2.65
		Craneopharyngioma	11	1.82
		Dermoid cyst	13	2.15
	WHO II	Low grade gliomas	112	18.54
		Meningioma	21	3.48
		Astrocytoma	10	1.67
		Ependymoma	7	1.16
		Frontal cavernoma	4	0.66
Malignant	WHO III	High grade gliomas	147	14.77
		Anaplastic astrocytomas	22	2.21
		Anaplastic meningioma	11	1.10
		Anaplastic oligodendro glioma	29	2.91
		Central astrocytomas	65	6.53
	WHO IV	Glioblastomas	442	44.42
		Medulloblastoma	109	10.95
		Metasasis	170	17.08

According to the radiological and histopathological reports, there were 995 malignant brain image slices, including WHO (World Health Organization) Grade IV tumors; 442 glioblastomas, 109 medulloblastoma, 170 metasasis/residual malignancies, and WHO Grade III tumors; 147 high grade gliomas, 22 anaplastic astrocytomas, 11 anaplastic meningioma, 29 anaplastic oligodendro glioma, 65 central astrocytomas within the population. Also there were 604 benign brain tumors slices with WHO Grade I; 13 pilocytic astrocytoma, 262 meningioma, 135 shwannoma, 16 hemangioblastoma, 11 craneopharyngioma, 13 Dermoid cysts, and WHO Grade II; 10 astrocytoma, 21 meningiomas, 112 low grade gliomas, 7 ependymomas, 4 frontal cavernoma

ISSN: 1475-925X