convert coordinates of deep feature patches to image-based physical coordinates

High-level function for handling the converion between voxel and corner-based indices of patches to center and physical-based landmark coordinates.

convertPatchCoordsToSpatialPoints(patchCoords, img, patchSize = 32)

Arguments

patchCoords	input patch coordinates (usually from deepFeatures)
img	reference image defining spatial domain
patchSize	vector or scalar defining patch dimensions

Value

spatial patch coordinates

Author

Avants BB

Examples

library(ANTsR)
img <- ri( 1 ) %>% iMath( "Normalize" )
mask = randomMask( getMask( img ), 20 )
features = deepFeatures( img, mask, patchSize = 32 )
#> Loaded Tensorflow version 2.6.0
coords = convertPatchCoordsToSpatialPoints( features$patchCoords, img )