import numpy as np
import nibabel as nb
import os
import sys
import nighresjava
from ..io import load_volume, save_volume
from ..utils import _output_dir_4saving, _fname_4saving, \
_check_topology_lut_dir, _check_available_memory
[docs]def background_estimation(image, distribution='exponential', ratio=1e-3,
skip_zero=True, iterate=True, dilate=0,
threshold=0.5,
save_data=False, overwrite=False, output_dir=None,
file_name=None):
""" Background Estimation
Estimates image background by robustlyfitting various distribution models
Parameters
----------
image: niimg
Input image
distribution: {'exponential','half-normal'}
Distribution model to use for the background noise
ratio: float, optional
Robustness ratio for estimating image intensities
skip_zero: bool, optional
Whether to consider or skip zero values
iterate: bool, optional
Whether to run an iterative estimation (preferred, but sometimes unstable)
dilate: int, optional
Number of voxels to dilate or erode in the final mask
save_data: bool
Save output data to file (default is False)
overwrite: bool
Overwrite existing results (default is False)
output_dir: str, optional
Path to desired output directory, will be created if it doesn't exist
file_name: str, optional
Desired base name for output files with file extension
(suffixes will be added)
Returns
----------
dict
Dictionary collecting outputs under the following keys
(suffix of output files in brackets)
* masked (niimg): The background-masked input image
* proba (niimg): The probability map of the foreground
* mask (niimg): The mask of the foreground
Notes
----------
Original Java module by Pierre-Louis Bazin.
"""
print('\nBackground Estimation')
# make sure that saving related parameters are correct
if save_data:
output_dir = _output_dir_4saving(output_dir, image)
masked_file = os.path.join(output_dir,
_fname_4saving(file_name=file_name,
rootfile=image,
suffix='bge-masked'))
proba_file = os.path.join(output_dir,
_fname_4saving(file_name=file_name,
rootfile=image,
suffix='bge-proba'))
mask_file = os.path.join(output_dir,
_fname_4saving(file_name=file_name,
rootfile=image,
suffix='bge-mask'))
if overwrite is False \
and os.path.isfile(masked_file) and os.path.isfile(proba_file) \
and os.path.isfile(mask_file) :
print("skip computation (use existing results)")
output = {'masked': load_volume(masked_file),
'proba': load_volume(proba_file),
'mask': load_volume(mask_file)}
return output
# start virtual machine, if not already running
try:
mem = _check_available_memory()
nighresjava.initVM(initialheap=mem['init'], maxheap=mem['max'])
except ValueError:
pass
# create instance
bge = nighresjava.IntensityBackgroundEstimator2()
# set parameters
# load image and use it to set dimensions and resolution
img = load_volume(image)
data = img.get_data()
affine = img.affine
header = img.header
resolution = [x.item() for x in header.get_zooms()]
dimensions = data.shape
if len(dimensions)>2:
bge.setDimensions(dimensions[0], dimensions[1], dimensions[2])
#bge.setResolutions(resolution[0], resolution[1], resolution[2])
else:
bge.setDimensions(dimensions[0], dimensions[1], 1)
#bge.setResolutions(resolution[0], resolution[1], 1)
bge.setInputImage(nighresjava.JArray('float')(
(data.flatten('F')).astype(float)))
# set algorithm parameters
bge.setBackgroundDistribution(distribution)
bge.setRobustMinMaxThresholding(ratio)
bge.setSkipZeroValues(skip_zero)
bge.setIterative(iterate)
bge.setDilateMask(dilate)
bge.setMaskThreshold(threshold)
# execute the algorithm
try:
bge.execute()
except:
# if the Java module fails, reraise the error it throws
print("\n The underlying Java code did not execute cleanly: ")
print(sys.exc_info()[0])
raise
return
# reshape output to what nibabel likes
masked_data = np.reshape(np.array(bge.getMaskedImage(),
dtype=np.float32), dimensions, 'F')
# adapt header max for each image so that correct max is displayed
# and create nifiti objects
header['cal_min'] = np.nanmin(masked_data)
header['cal_max'] = np.nanmax(masked_data)
masked = nb.Nifti1Image(masked_data, affine, header)
proba_data = np.reshape(np.array(bge.getProbaImage(),
dtype=np.float32), dimensions, 'F')
# adapt header max for each image so that correct max is displayed
# and create nifiti objects
header['cal_min'] = np.nanmin(proba_data)
header['cal_max'] = np.nanmax(proba_data)
proba = nb.Nifti1Image(proba_data, affine, header)
mask_data = np.reshape(np.array(bge.getMask(),
dtype=np.float32), dimensions, 'F')
# adapt header max for each image so that correct max is displayed
# and create nifiti objects
header['cal_min'] = np.nanmin(mask_data)
header['cal_max'] = np.nanmax(mask_data)
mask = nb.Nifti1Image(mask_data, affine, header)
if save_data:
save_volume(masked_file, masked)
save_volume(proba_file, proba)
save_volume(mask_file, mask)
return {'masked': masked, 'proba': proba, 'mask': mask}