Skip to main content

Table 1 Summary of feature groups in proposed Radiomics-Driven Statistical Textural Distinctiveness (RD-STD) [14]

From: MPCaD: a multi-scale radiomics-driven framework for automated prostate cancer localization and detection

Feature group

Number of features

Description

Textural (1st-order)

4

Mean, Standard deviation, Kurtosis, Skewness

  

Energy, contrast, correlation, variance, inverse difference moment normalized,

Textural (2nd-order)

72

Sum average, sum variance, entropy, sum entropy, difference entropy,

 

(18 in each of 4 directions)

Information measure of correlation, homogeneity, autocorrelation

  

Difference variance, dissimilarity, cluster shade, cluster prominence, maximum probability

Gabor filters

12

3 scales and 4 orientations

Kirsch filters

8

8 directions

Total

96

All features