Classification of Region of Interest in Mammograms Using Dual Contourlet Transform and Improved KNN

Table 7

Classification performance of different methods.

Reference/year	Database	Classification technique		Accuracy (%)	AUC (Az)
Proposed method	MIAS	Improved KNN classifier	N versus A	94.14	0.9582
			B versus M	95.76	0.9717
Ref. [16] 2015	DDSM; MIAS	Random forest classifier	B versus M	91.73	0.9467
Ref. [5] 2014	MIAS	SVM classifier	N versus A	85.48	—
Ref. [34] 2013	MIAS	OWBPE classifier.	B versus M	89.28	0.9280
Ref. [7] 2012	MIAS	SVM classifier	N versus A	95.98	—
Ref. [6] 2011	MIAS	MLP classifier	B versus M	93.60	—
Ref. [35] 2010	DDSM	MLP classifier	B versus M	88.75	—
Ref. [33] 2008	MIAS	S2SP	B versus M	—	0.95

N versus A represents normal versus abnormal; B versus M represents benign versus malignant.