Skip to main content

Table 1 Features extracted from each image and time-point of the series

From: A machine learning approach for differentiating malignant from benign enhancing foci on breast MRI

 

Feature class

Intensity featuresa

Shape featuresb

GLCM featuresc (mean and standard deviation computed along the 3D directions)

GLRLM featuresd (mean and standard deviation computed along the 3D directions)

Features

Max

Eccentricity

Energy

Short run emphasis

Min

Elongation

Entropy

Long run emphasis

Mean

Major axis length (mm)

Inversion different moment

Grey level non-uniformity

Sigma

Minor axis length (mm)

Inertia

Run length non-uniformity

Variance

Volume (mm3)

Cluster shade

Low grey level run emphasis

Integrated intensity

 

Cluster prominence

High grey level run emphasis

  

Short run low grey level emphasis

Short run high grey level emphasis

Long run low grey level emphasis

Long run high grey level emphasis

  1. 3D three-dimensional
  2. aIntensity features: first order statistics calculated from pixel intensities
  3. bShape features: 3D shape descriptors
  4. cGrey level co-occurrence matrix features: they are reported as average and standard deviation computed along all the three-dimensional directions
  5. dGrey level run length matrix features: they are reported as average and standard deviation computed along all the 3D directions