Skip to main content
Fig. 1 | Visual Computing for Industry, Biomedicine, and Art

Fig. 1

From: Robustness of radiomic features in magnetic resonance imaging: review and a phantom study

Fig. 1

Schematic representation of workflow in this study. Image segmentation is performed manually on a single image. The ROIs are interpolated to images of different in-plane resolutions for voxel size analysis. Gaussian noise is added to generate different signal to noise ratio steps and generate 10 different noise realizations for test-retest analysis. Shape, first order, GLCM and GLRLM features are calculated for each ROI. GLCLM and GLRLM features are calculated after normalization (mean ± 3SD or zero to maximum) and discretization (64 gray levels). ROI Region of interest, GLCM Gray level cooccurrence matrix, GLRLM Gray level run length matrix

Back to article page