Skip to main content
. 2010 May 28;24(3):446–463. doi: 10.1007/s10278-010-9298-1

Table 5.

Results of Support Vector Machine Classifier for Top-Performing Individual Attributes from Each Feature Class in Distinguishing Benign from Malignant Lesions Using Leave-One-Out Validation

Feature class Feature Inline graphic Inline graphic Inline graphic Inline graphic
Morphological Smoothness 0.73 0.88 0.53 0.77
Precontrast texture Gabor filter Inline graphic 0.63 0.90 0.25 0.65
Postcontrast texture Intensity variance 0.68 0.83 0.47 0.70
Signal intensity Signal intensity kinetics 0.63 0.67 0.59 0.75
First-order textural kinetics Gabor filter Inline graphic 0.76 0.75 0.76 0.78
Second-order textural kinetics Contrast inverse moment 0.73 0.88 0.52 0.70