fir_utils.py

def maximum_idx(array):
    """finds the index of the center-of-mass"""
    import numpy as np
    import sys

    if not isinstance(array,np.ndarray):
        array = np.array(array)

    return np.unravel_index( array.argmax(axis=None), array.shape)

def functional_binarize(array,bool_func):
    """binarizes the array by settting values that bool_func 
    returns true to 1. Bool func must accept array as its first 
    input"""
    import numpy as np

    binarized_array = np.zeros(array.shape)

    binarized_array[bool_func(array)] = 1

    return binarized_array

def is_binary(array):
    import numpy as np
    return (np.unique(array) == np.array([0,1])).all()

def binarize_peak(array):
    """ finds the peak value and binarizes it"""
    import numpy as np
    
    binarized_peak = np.zeros(array.shape)
    binarized_peak[maximum_idx(array)] = 1

    return binarized_peak
    #return functional_binarize(np.array(array),lambda a: a == a[maximum_idx(a)])

def dilate_mask(mask, iterations=1):
    """dilates the mask iterations number of times"""
    from scipy.ndimage.morphology import binary_dilation
    import sys

    if not is_binary(mask):
        sys.exit("MASK TO BE DILATED WAS NOT A BINARY MASK")

    mask[binary_dilation(mask,iterations=iterations)] = 1
    return mask

def convert_affine(unwarped_brain, mean_func, out_fsl_file):
    """Converts fsl-style Affine registration into ANTS compatible itk format

    Parameters
    ----------
    unwarped_brain : structural reference image
    mean_func : image that was coregistered
    out_fsl_file : fsl-style coregistration matrix

    Returns
    -------
    file : returns the filename corresponding to the converted registration
    """
    import os
    cmd = "c3d_affine_tool -ref %s -src %s %s -fsl2ras \
-oitk fsl2antsAffine.txt" % (unwarped_brain, mean_func, out_fsl_file)
    os.system(cmd)
    return os.path.abspath('fsl2antsAffine.txt')