nx10 / fineslice

Sample 3D-affine transformed images

fineslice

fineslice is a lightweight sampler for 3D-affine transformed images (commonly used in neuroscience) implemented in pure Python + NumPy.

It does not make any assumptions about the data. Pass any image texture and affine matrix directly into it.

Features

• Precision sampling (no need to 're-sample' and loose precision)
• Automatically finds optimal dimensions
• Only depends on NumPy

Usage with nibabel

For the best performance directly pass in the nibabel data object as a texture:

import nibabel as nib
import fineslice as fine

img = nib.load('my_image.nii.gz')

out = fine.sample_0d(
    texture=img.dataobj,
    affine=img.affine,
    out_position=(0, 0, 0)
)