Research Article
Differentiation of Pelvic Osteosarcoma and Ewing Sarcoma Using Radiomic Analysis Based on T2-Weighted Images and Contrast-Enhanced T1-Weighted Images
(a) Numbers of radiomic feature selection by different methods |
| | T2-FS | CET1 |
| | Independent-sample -test | 141 | 60 | | Spearman test | 27 | 10 | | LASSO | 9 | 7 |
|
|
(b) Radiomic feature selection by the LASSO method |
| | No. | T2-FS | CET1 |
| | 1 | Intercept | Intercept | | 2 | Correlation_All Direction_offset7_SD | Min intensity | | 3 | Surface volume ratio | Inverse Difference Moment_All Direction_offset 7_SD | | 4 | GLCM Energy_All Direction_offset4_SD | GLCM Entropy_All Direction_offset 1_SD | | 5 | Inverse Difference Moment_All Direction_offset 4_SD | GLCM Energy_angle135_offset 7 | | 6 | Inverse Difference Moment_All Direction_offset 7_SD | Volume MM | | 7 | Cluster Prominence_All Direction_offset 7_SD | Surface volume ratio | | 8 | High Grey Level Run Emphasis_All Direction_offset 7_SD | | | 9 | Short Run High Grey Level Emphasis_All Direction_offset 7_SD | |
|
|
Note: GLCM = gray-level cooccurrence matrix.
|
