Table 2 Feature selection: most important Radiomics features (in descending order of importance).
From: Pseudoprogression prediction in high grade primary CNS tumors by use of radiomics
Level of importance | Feature |
|---|---|
1 | orig.ngtdm.Strength |
2 | Age |
3 | orig.glcm.ClusterShade |
4 | orig.shape.MinorAxisLength |
5 | orig.shape.Elongation |
6 | orig.glrlm.LongRunHighGrayLevelEmphasis |
7 | orig.ngtdm.Busyness |
8 | orig.shape.Sphericity |
9 | orig.glcm.Imc2 |
10 | orig.glszm.LowGrayLevelZoneEmphasis |
11 | orig.glcm.MCC |
12 | orig.fst.ord.RobustMeanAbsoluteDeviation |
13 | orig.fst.ord.Median |
14 | orig.gldm.LowGrayLevelEmphasis |
15 | orig.ngtdm.Contrast |
