Skip to main content

Table 1 Summary of textural features used in the feature model

From: Automated prostate cancer detection via comprehensive multi-parametric magnetic resonance imaging texture feature models

Feature class Feature
First-order statistical features Mean, Standard deviation
  Skewness
  Kurtosis
Second-order statistical Energy, Contrast
features (Haralick) Correlation, Variance
  Inverse difference moment
  Sum average, Sum variance
  Sum entropy, Entropy
  Difference variance
  Difference entropy
  Information measure of correlation
  Homogeneity, Autocorrelation
  Dissimilarity, Cluster shade
  Cluster prominence
  Maximum probability
Gabor filters 3 scales and 4 orientations
Kirsch filters 8 directions