A Computer-Aided Diagnosis System for Dynamic Contrast-Enhanced MR Images Based on Level Set Segmentation and ReliefF Feature Selection

Table 3

Classification results of the different segmentation methods (leave-one-case-out).

Segmentation method	Classification model	Accuracy (%)	Sensitivity (%)	Specificity (%)
FCM	Fisher	83.3	91.0	69.0
	SVM	80.8	93.6	57.1
FCM-GVF-snake	Fisher	78.3	82.1	71.4
	SVM	82.5	93.6	61.9
CV-level set	Fisher	90.8	94.9	83.3
	SVM	90.0	98.7	73.8