Skip to main content
European Radiology Experimental

Table 4 Performance of different feature combinations using the random forest method

From: Performance of machine learning software to classify breast lesions using BI-RADS radiomic features on ultrasound images

Features

Sensitivity (%)

Specificity (%)

Area under the curve

{4}

62.2

71.7

0.697

{4, 8}

72.3

74.6

0.760

{4, 8, 5}

72.6

72.6

0.778

{4, 8, 5, 2}

72.7

75.9

0.811

  1. {2}: orientation; {4}: number of peaks on the distance vector (NumPeaks); {5}: average of distance vector; {8}: entropy
Back to article page