Motivations

An neurological or anatomical image is usually a file saved in nifti format. The algorithm written to manipulate it can act on the image header or on the image data (or the corresponding numpy array) where is it tested.

Apply the algorithm to the image usually involves the repetitive task of loading the image with nibabel, extract the header or the data, apply the algorithm, create a new instance of the modified nifti image, and finally save it.

LABelsToolkit is aimed at reducing this at one step, allowing to have as input and output directly the path to the input image and the output image.