unravel.register.reg_prep module#

Use reg_prep (rp) from UNRAVEL to load a full resolution autofluo image and resamples to a lower resolution for registration.

Input examples (path is relative to ./sample??; 1st glob match processed):

*.czi, autofluo/*.tif series, autofluo, *.tif, or *.h5

Outputs:

./sample??/reg_inputs/autofl_`*`um.nii.gz ./sample??/reg_inputs/autofl_`*`um_tifs/*.tif series (used for training ilastik for seg_brain_mask)

Note

  • If -d is not provided, the current directory is used to search for sample?? dirs to process.

  • If the current dir is a sample?? dir, it will be processed.

  • If -d is provided, the specified dirs and/or dirs containing sample?? dirs will be processed.

  • If -p is not provided, the default pattern for dirs to process is ‘sample??’.

Next command:

seg_copy_tifs for seg_brain_mask or reg

Usage:#

reg_prep -i *.czi [-md path/metadata.txt] [For .czi: –channel 0] [-o reg_inputs/autofl_50um.nii.gz] [–reg_res 50] [–zoom_order 0] [–miracl] [-d list of paths] [-p sample??] [-v]

unravel.register.reg_prep.parse_args()[source]#
unravel.register.reg_prep.reg_prep(ndarray, xy_res, z_res, reg_res, zoom_order, miracl)[source]#

Prepare the autofluo image for reg or mimic preprocessing for vstats_prep.

Parameters:

  • ndarray (-) – full res 3D autofluo image.

  • xy_res (-) – x/y resolution in microns of ndarray.

  • z_res (-) – z resolution in microns of ndarray.

  • reg_res (-) – Resample input to this resolution in microns for reg.

  • zoom_order (-) – Order for resampling (scipy.ndimage.zoom).

  • miracl (-) – Include reorientation step to mimic MIRACL’s tif to .nii.gz conversion.

Returns:

Resampled image.

Return type:

  • img_resampled (np.ndarray)

unravel.register.reg_prep.main()[source]#