jrenslo
jrenslo
Minor suggestion, but for completeness sake it would be helpful to know what dimension and format of images are supported, and what the expected format of the "mask" image should...
Typically multiple acquisitions are stacked in another dimension (e.g. 512x512x128x2) but default behavior of rt-utils returns the slices in a single stacked axis. (512x512x256). https://github.com/qurit/rt-utils/blob/8ba5cf64bd0ea5119523da0bafd10174115eebe4/rt_utils/image_helper.py#L239
I'm not certain, but when a stack of DICOM has nonuniform slice thickness, creation of a mask may be incorrect due to application of the transformation matrix to the stacked...