Calculate radiomic features on the partitioned 3D lung
radiomics_partition( img, mask, sides = c("right", "left"), featuresFirst = c("mean", "sd", "skew", "kurtosis", "min", "q1", "median", "q3", "max", "energy", "rms", "uniformity", "entropy"), featuresSpatial = c("mi", "gc", "fd"), partition = NULL, kernel_size = c(30, 30, 30), kernel_stride = c(30, 30, 30), threshold = 1000, tidy = TRUE )
| img | CT scan in ANTs image file format |
|---|---|
| mask | Mask of CT scan in ANTs image file format |
| sides | Choose to calculate radiomic features on the right and/or left lungs. Note: Right lung = 1, left lung = 2, non-lung = 0 |
| featuresFirst | First level radiomic features to calculate |
| featuresSpatial | Spatial radiomic features to calculate |
| partition | Matrix of x, y, and z coordinates for each partition from partition_lung. If null, partition_lung is called. |
| kernel_size | (If partition is null) Size of the kernel, in voxel units of width, depth, and height. Must be c(3,3,3) or greater. Default: c(30,30,30) |
| kernel_stride | (If partition is null) Stride (or spacing) between kernels, in voxel units, for width, depth, and height. If kernel_stride = kernel_size, the partitions are non-overlapping. If stride = c(1,1,1), then each voxel is returned. |
| threshold | Number of non-missing voxels needed to calculate radiomic features in each partition. |
| tidy | Logical. If true, outputs a tidy dataframe with results. If false, outputs nested loop. |
Values from selected features for both left and right lungs