Table 2.
Details of the extracted feature sets used in the two-stage renal tumor classification.
Texture Features | |
Feature Set 1: First-order (histogram features) | 6 features |
Feature Set 2: First-order (percentiles) | 10 features |
Feature Set 3: Second-order (GLCM) | 6 features |
Shape Features | |
Feature Set 4: Spherical harmonic reconstruction errors | 70 features |
Functional Features | |
Feature Set 5: Wash-in/out slopes | 2 features |
Combined Features | |
Feature Sets 1, 2, 3, 4, and 5 | 94 features |