Skip to main content
Visual Computing for Industry, Biomedicine, and Art

Table 6 Number of features of high, moderate and low robustness in each feature class, as defined by average of intraclass correlation coefficient over 10 noise realizations in reference to pixel size with normalization of mean ± 3SD or zero to maximum

From: Robustness of radiomic features in magnetic resonance imaging: review and a phantom study

Feature group

High (ICC > 0.9)

Moderate (ICC 0.5–0.9)

Low (ICC < 0.5)

 

Mean ± 3SD

Zero to maximum

Mean ± 3SD

Zero to maximum

Mean ± 3SD

Zero to maximum

First order

8/13

8/13

4/13

4/13

1/13

1/13

Shape

10/10

10/10

0/10

0/10

0/10

0/10

GLCM

12/22

19/22

9/22

3/22

1/22

0/22

GLRLM

6/11

10/11

3/11

1/11

2/11

0/11

  1. The denominator in the table signifies the total number of features in the feature class (i.e., first order, shape, GLCM or GLRLM). GLCM Gray level cooccurrence matrix, GLRLM Gray level run length matrix, ICC Intraclass correlation coefficient
Back to article page