apply_mrtm2_to_all_voxels#

petpal.kinetic_modeling.parametric_images.apply_mrtm2_to_all_voxels(tac_times_in_minutes: numpy.ndarray, tgt_image: numpy.ndarray, ref_tac_vals: numpy.ndarray, k2_prime: float, t_thresh_in_mins: float, mask_img: numpy.ndarray) → Tuple[numpy.ndarray, numpy.ndarray]#

Generates parametric images for 4D-PET data using the MRTM2 reference tissue method.

Parameters:

  • tac_times_in_minutes (np.ndarray) – A 1D array representing the reference TAC and PET frame times in minutes.

  • tgt_image (np.ndarray) – A 4D array representing the 3D PET image over time. The shape of this array should be (x, y, z, time).

  • ref_tac_vals (np.ndarray) – A 1D array representing the reference TAC values. This array should be of the same length as tac_times_in_minutes.

  • k2_prime (float) – A float representing the k2’ value to be used for MRTM2 analysis. This is chosen based on the tracer or based on a regional MRTM1 analysis.

  • t_thresh_in_mins (float) – A float representing the threshold time past which MRTM parameters are calculated with a least squares fit.

  • mask_img (np.ndarray) – A 3D array representing the brain mask for tgt_image, where brain regions are labelled 1 and non-brain regions are labelled 0. This is made necessary in order to save time during computation.

Returns:

bp_img (np.ndarray) –

A 3D array with computed BP values based on the MRTM2 parameter fit

results.

simulation_img (np.ndarray): A 4D array with the same shape as tgt_image where each voxel

is the best fit curve based on the solved parameters to the linear equation in MRTM2.

On this page

This Page