A Benign and Malignant Breast Tumor Classification Method via Efficiently Combining Texture and Morphological Features on Ultrasound Images

Table 2

The classification results based on the methods of single features with single classifier.

Method		Evaluation (%)
Feature	Classifier	Accuracy	Sensitivity	Specificity
LBP	SVM [17]	85.56	86.79	83.78
	KNN [25]	84.44	84.91	83.78
	DT [26]	66.33	58.49	81.08
	LDA [27]	74.44	77.36	70.27
HOG	SVM [17]	81.11	84.91	75.68
	KNN [25]	61.11	100.00	5.41
	DT [26]	67.78	67.92	67.57
	LDA [27]	70.00	75.47	62.16
GLCM	SVM [17]	78.89	92.45	59.46
	KNN [25]	65.56	75.47	51.35
	DT [26]	71.11	77.36	62.16
	LDA [27]	74.44	84.91	59.46
LBP+HOG+GLCM	SVM [17]	86.67	92.45	78.38
	KNN [25]	64.44	100.00	13.51
	DT [26]	72.22	73.58	70.27
	LDA [27]	75.56	84.91	62.16
Morphological	SVM [17]	75.56	67.92	86.49
	NB [18]	81.11	69.81	97.30
	LDA [26]	75.56	60.38	97.30