HIGHLIGHTS
SUMMARY
It has been observed that radiomic features are dependent on image perturbation; moreover, radiomic features reproducibility might depend on region-of-interest (ROI) segmentation performed by different physicians. The authors evaluated the robustness of radiomic features across segmentations and evaluated the influence of other parameters (specifically, bin-width, pixel-distance in 2 of 16 second-order features, interpolator and isotropic resolution) on a set of CT volumes from patients suffering from NSCLC. To assess the dependence on important parameters used for features computation, the authors computed features using several values of bin-width and . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.