Table 2.
The number of extracted features from different feature extractor tools feature families, types, and groups. The fifth column indicates whether features were extracted from two-dimensional images (per slice) or image volumes. Finally, the last column indicates the per-pixel or per-image feature extraction approach
| Origin of code | Feature type | Feature group | Number of features | 2D/3D | Per-pixel/per-image |
|---|---|---|---|---|---|
| SERA package | Morphology | - | 29 | 3D | Per-image |
| Textural | GLCM | 50 |
3D and 2D |
||
| GLRLM | 32 | ||||
| GLSZM | 32 | ||||
| GLDZM | 32 | ||||
| NGTDM | 10 | ||||
| NGLDM | 34 | ||||
|
First-order Intensity-based |
- | 50 | - | Per-image | |
| Self-generated | Textural |
GLCM/ Haralick |
14 | 2D | Per-pixel |
| LAWS | 25 | Per-pixel | |||
| Transform-based | Gabor | 48 | Per-image |