apply_mask_4d#

petpal.preproc.regional_tac_extraction.apply_mask_4d(input_arr: numpy.ndarray, mask_arr: numpy.ndarray, verbose: bool = False) → numpy.ndarray#

Function to extract ROI voxel tacs from an array using a mask array.

This function applies a 3D mask to a 4D image, returning the time series for each voxel in a single flattened numpy array.

Parameters:

  • input_arr (np.ndarray) – Input 4D-image from which to extract ROI voxel tacs.

  • mask_arr (np.ndarray) – Mask image which determines which voxels to extract.

  • verbose (bool, optional) – If True, prints information about the shape of extracted voxel tacs.

Returns:

out_voxels (np.ndarray) – Time series of each voxel in the mask, as a flattened numpy array.

Raises:

  • AssertionError – If input array is not 4D.

  • AssertionError – If input and mask array shapes are mismatched.

Example

import ants
import numpy as np

from petpal.preproc.regional_tac_extraction import apply_mask_4d

# Read images
pet_img = ants.image_read("/path/to/pet.nii.gz")
masked_region_img = ants.image_read("/path/to/mask_region.nii.gz")

# Get underlying arrays
pet_arr = pet_img.numpy()
masked_region_arr = masked_region_img.numpy()

# Run ROI extraction and save
time_series = apply_mask_4d(input_arr=pet_arr, mask_arr=masked_region_arr).T
np.savetxt("time_series.tsv", time_series, delimiter='      ')
On this page

This Page