Research Article
Diagnostic Accuracy of Machine Learning-Based Radiomics in Grading Gliomas: Systematic Review and Meta-Analysis
Table 4
Summary of the methods used in the reviewed studies.
| Study and year | Data source | External validation | Feature type | Feature extraction | Feature selection | Segmentation |
| Cho et al. 2018 | Public | Training + testing | First-order and second-order (GLCM, ISZ) | 486 | 5 | ROI | Tian et al. 2018 | Private | Training | First-order, second-order (GLCM, GLCGM) | 510 | 30 | VOI | Hashido et al. 2018 | Private | Training (42) + testing (4) | First-order, second-order (GLCM, GLDM, GLRLM, GLSZM, and NGTDM) | 91 | 75 | Random forest-based semiautomatic tumor segmentation | Vamvakas et al. 2019 | Private | Training | First-order and second-order texture (GLCM, GLRLM) | 581 | 21 | VOI | Zhao et al. 2020 | Private | Training | First-order and second-order (GLCM, GLRLM, GLSZM, and GLDM) | 1072 | 30 | VOI |
|
|