def threshold_map(maps, mask, ref_img, threshold, csize=None):
"""Perform cluster-extent thresholding.
Parameters
----------
maps : (M x C) array_like
Statistical maps to be thresholded.
mask : (S) array_like
Binary mask.
ref_img : img_like
Reference image to convert to niimgs with.
threshold : :obj:`float`
Value threshold to apply to maps.
csize : :obj:`int` or :obj:`None`, optional
Minimum cluster size. If None, standard thresholding (non-cluster-extent) will be done.
Default is None.
Returns
-------
maps_thresh : (M x C) array_like
"""
n_voxels, n_components = maps.shape
maps_thresh = np.zeros([n_voxels, n_components], bool)
if csize is None:
csize = np.max([int(n_voxels * 0.0005) + 5, 20])
else:
csize = int(csize)
for i_comp in range(n_components):
# Cluster-extent threshold and binarize F-maps
ccimg = io.new_nii_like(
ref_img, np.squeeze(utils.unmask(maps[:, i_comp], mask)))
maps_thresh[:, i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=threshold,
mask=mask,
binarize=True)
return maps_thresh
def dependence_metrics(catd,
tsoc,
mmix,
t2s,
tes,
ref_img,
reindex=False,
mmixN=None,
algorithm=None,
label=None,
out_dir='.',
verbose=False):
"""
Fit TE-dependence and -independence models to components.
Parameters
----------
catd : (S x E x T) array_like
Input data, where `S` is samples, `E` is echos, and `T` is time
tsoc : (S x T) array_like
Optimally combined data
mmix : (T x C) array_like
Mixing matrix for converting input data to component space, where `C`
is components and `T` is the same as in `catd`
t2s : (S [x T]) array_like
Limited T2* map or timeseries.
tes : list
List of echo times associated with `catd`, in milliseconds
ref_img : str or img_like
Reference image to dictate how outputs are saved to disk
reindex : bool, optional
Whether to sort components in descending order by Kappa. Default: False
mmixN : (T x C) array_like, optional
Z-scored mixing matrix. Default: None
algorithm : {'kundu_v2', 'kundu_v3', None}, optional
Decision tree to be applied to metrics. Determines which maps will be
generated and stored in seldict. Default: None
label : :obj:`str` or None, optional
Prefix to apply to generated files. Default is None.
out_dir : :obj:`str`, optional
Output directory for generated files. Default is current working
directory.
verbose : :obj:`bool`, optional
Whether or not to generate additional files. Default is False.
Returns
-------
comptable : (C x X) :obj:`pandas.DataFrame`
Component metric table. One row for each component, with a column for
each metric. The index is the component number.
seldict : :obj:`dict` or None
Dictionary containing component-specific metric maps to be used for
component selection. If `algorithm` is None, then seldict will be None as
well.
betas : :obj:`numpy.ndarray`
mmix_new : :obj:`numpy.ndarray`
"""
# Use t2s as mask
mask = t2s != 0
if not (catd.shape[0] == t2s.shape[0] == mask.shape[0] == tsoc.shape[0]):
raise ValueError('First dimensions (number of samples) of catd ({0}), '
'tsoc ({1}), and t2s ({2}) do not '
'match'.format(catd.shape[0], tsoc.shape[0],
t2s.shape[0]))
elif catd.shape[1] != len(tes):
raise ValueError('Second dimension of catd ({0}) does not match '
'number of echoes provided (tes; '
'{1})'.format(catd.shape[1], len(tes)))
elif not (catd.shape[2] == tsoc.shape[1] == mmix.shape[0]):
raise ValueError('Number of volumes in catd ({0}), '
'tsoc ({1}), and mmix ({2}) do not '
'match.'.format(catd.shape[2], tsoc.shape[1],
mmix.shape[0]))
elif t2s.ndim == 2:
if catd.shape[2] != t2s.shape[1]:
raise ValueError('Number of volumes in catd '
'({0}) does not match number of volumes in '
't2s ({1})'.format(catd.shape[2], t2s.shape[1]))
# mask everything we can
tsoc = tsoc[mask, :]
catd = catd[mask, ...]
t2s = t2s[mask]
# demean optimal combination
tsoc_dm = tsoc - tsoc.mean(axis=-1, keepdims=True)
# compute un-normalized weight dataset (features)
if mmixN is None:
mmixN = mmix
WTS = computefeats2(tsoc, mmixN, mask=None, normalize=False)
# compute PSC dataset - shouldn't have to refit data
tsoc_B = get_coeffs(tsoc_dm, mmix, mask=None)
del tsoc_dm
tsoc_Babs = np.abs(tsoc_B)
PSC = tsoc_B / tsoc.mean(axis=-1, keepdims=True) * 100
# compute skews to determine signs based on unnormalized weights,
# correct mmix & WTS signs based on spatial distribution tails
signs = stats.skew(WTS, axis=0)
signs /= np.abs(signs)
mmix = mmix.copy()
mmix *= signs
WTS *= signs
PSC *= signs
totvar = (tsoc_B**2).sum()
totvar_norm = (WTS**2).sum()
# compute Betas and means over TEs for TE-dependence analysis
betas = get_coeffs(utils.unmask(catd, mask), mmix,
np.repeat(mask[:, np.newaxis], len(tes), axis=1))
betas = betas[mask, ...]
n_voxels, n_echos, n_components = betas.shape
mu = catd.mean(axis=-1, dtype=float)
tes = np.reshape(tes, (n_echos, 1))
fmin, _, _ = getfbounds(n_echos)
# set up Xmats
X1 = mu.T # Model 1
X2 = np.tile(tes, (1, n_voxels)) * mu.T / t2s.T # Model 2
# tables for component selection
kappas = np.zeros([n_components])
rhos = np.zeros([n_components])
varex = np.zeros([n_components])
varex_norm = np.zeros([n_components])
Z_maps = np.zeros([n_voxels, n_components])
F_R2_maps = np.zeros([n_voxels, n_components])
F_S0_maps = np.zeros([n_voxels, n_components])
pred_R2_maps = np.zeros([n_voxels, n_echos, n_components])
pred_S0_maps = np.zeros([n_voxels, n_echos, n_components])
LGR.info('Fitting TE- and S0-dependent models to components')
for i_comp in range(n_components):
# size of comp_betas is (n_echoes, n_samples)
comp_betas = np.atleast_3d(betas)[:, :, i_comp].T
alpha = (np.abs(comp_betas)**2).sum(axis=0)
varex[i_comp] = (tsoc_B[:, i_comp]**2).sum() / totvar * 100.
varex_norm[i_comp] = (WTS[:, i_comp]**2).sum() / totvar_norm
# S0 Model
# (S,) model coefficient map
coeffs_S0 = (comp_betas * X1).sum(axis=0) / (X1**2).sum(axis=0)
pred_S0 = X1 * np.tile(coeffs_S0, (n_echos, 1))
pred_S0_maps[:, :, i_comp] = pred_S0.T
SSE_S0 = (comp_betas - pred_S0)**2
SSE_S0 = SSE_S0.sum(axis=0) # (S,) prediction error map
F_S0 = (alpha - SSE_S0) * (n_echos - 1) / (SSE_S0)
F_S0_maps[:, i_comp] = F_S0
# R2 Model
coeffs_R2 = (comp_betas * X2).sum(axis=0) / (X2**2).sum(axis=0)
pred_R2 = X2 * np.tile(coeffs_R2, (n_echos, 1))
pred_R2_maps[:, :, i_comp] = pred_R2.T
SSE_R2 = (comp_betas - pred_R2)**2
SSE_R2 = SSE_R2.sum(axis=0)
F_R2 = (alpha - SSE_R2) * (n_echos - 1) / (SSE_R2)
F_R2_maps[:, i_comp] = F_R2
# compute weights as Z-values
wtsZ = (WTS[:, i_comp] - WTS[:, i_comp].mean()) / WTS[:, i_comp].std()
wtsZ[np.abs(wtsZ) > Z_MAX] = (
Z_MAX * (np.abs(wtsZ) / wtsZ))[np.abs(wtsZ) > Z_MAX]
Z_maps[:, i_comp] = wtsZ
# compute Kappa and Rho
F_S0[F_S0 > F_MAX] = F_MAX
F_R2[F_R2 > F_MAX] = F_MAX
norm_weights = np.abs(wtsZ**2.)
kappas[i_comp] = np.average(F_R2, weights=norm_weights)
rhos[i_comp] = np.average(F_S0, weights=norm_weights)
del SSE_S0, SSE_R2, wtsZ, F_S0, F_R2, norm_weights, comp_betas
if algorithm != 'kundu_v3':
del WTS, PSC, tsoc_B
# tabulate component values
comptable = np.vstack([kappas, rhos, varex, varex_norm]).T
if reindex:
# re-index all components in descending Kappa order
sort_idx = comptable[:, 0].argsort()[::-1]
comptable = comptable[sort_idx, :]
mmix_new = mmix[:, sort_idx]
betas = betas[..., sort_idx]
pred_R2_maps = pred_R2_maps[:, :, sort_idx]
pred_S0_maps = pred_S0_maps[:, :, sort_idx]
F_R2_maps = F_R2_maps[:, sort_idx]
F_S0_maps = F_S0_maps[:, sort_idx]
Z_maps = Z_maps[:, sort_idx]
tsoc_Babs = tsoc_Babs[:, sort_idx]
if algorithm == 'kundu_v3':
WTS = WTS[:, sort_idx]
PSC = PSC[:, sort_idx]
tsoc_B = tsoc_B[:, sort_idx]
else:
mmix_new = mmix
del mmix
if verbose:
# Echo-specific weight maps for each of the ICA components.
io.filewrite(utils.unmask(betas, mask),
op.join(out_dir, '{0}betas_catd.nii'.format(label)),
ref_img)
# Echo-specific maps of predicted values for R2 and S0 models for each
# component.
io.filewrite(utils.unmask(pred_R2_maps, mask),
op.join(out_dir, '{0}R2_pred.nii'.format(label)), ref_img)
io.filewrite(utils.unmask(pred_S0_maps, mask),
op.join(out_dir, '{0}S0_pred.nii'.format(label)), ref_img)
# Weight maps used to average metrics across voxels
io.filewrite(utils.unmask(Z_maps**2., mask),
op.join(out_dir, '{0}metric_weights.nii'.format(label)),
ref_img)
del pred_R2_maps, pred_S0_maps
comptable = pd.DataFrame(comptable,
columns=[
'kappa', 'rho', 'variance explained',
'normalized variance explained'
])
comptable.index.name = 'component'
# Generate clustering criteria for component selection
if algorithm in ['kundu_v2', 'kundu_v3']:
Z_clmaps = np.zeros([n_voxels, n_components], bool)
F_R2_clmaps = np.zeros([n_voxels, n_components], bool)
F_S0_clmaps = np.zeros([n_voxels, n_components], bool)
Br_R2_clmaps = np.zeros([n_voxels, n_components], bool)
Br_S0_clmaps = np.zeros([n_voxels, n_components], bool)
LGR.info('Performing spatial clustering of components')
csize = np.max([int(n_voxels * 0.0005) + 5, 20])
LGR.debug('Using minimum cluster size: {}'.format(csize))
for i_comp in range(n_components):
# Cluster-extent threshold and binarize F-maps
ccimg = io.new_nii_like(
ref_img, np.squeeze(utils.unmask(F_R2_maps[:, i_comp], mask)))
F_R2_clmaps[:,
i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=fmin,
mask=mask,
binarize=True)
countsigFR2 = F_R2_clmaps[:, i_comp].sum()
ccimg = io.new_nii_like(
ref_img, np.squeeze(utils.unmask(F_S0_maps[:, i_comp], mask)))
F_S0_clmaps[:,
i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=fmin,
mask=mask,
binarize=True)
countsigFS0 = F_S0_clmaps[:, i_comp].sum()
# Cluster-extent threshold and binarize Z-maps with CDT of p < 0.05
ccimg = io.new_nii_like(
ref_img, np.squeeze(utils.unmask(Z_maps[:, i_comp], mask)))
Z_clmaps[:, i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=1.95,
mask=mask,
binarize=True)
# Cluster-extent threshold and binarize ranked signal-change map
ccimg = io.new_nii_like(
ref_img,
utils.unmask(stats.rankdata(tsoc_Babs[:, i_comp]), mask))
Br_R2_clmaps[:, i_comp] = utils.threshold_map(
ccimg,
min_cluster_size=csize,
threshold=(max(tsoc_Babs.shape) - countsigFR2),
mask=mask,
binarize=True)
Br_S0_clmaps[:, i_comp] = utils.threshold_map(
ccimg,
min_cluster_size=csize,
threshold=(max(tsoc_Babs.shape) - countsigFS0),
mask=mask,
binarize=True)
del ccimg, tsoc_Babs
if algorithm == 'kundu_v2':
# WTS, tsoc_B, PSC, and F_S0_maps are not used by Kundu v2.5
selvars = [
'Z_maps', 'F_R2_maps', 'Z_clmaps', 'F_R2_clmaps',
'F_S0_clmaps', 'Br_R2_clmaps', 'Br_S0_clmaps'
]
elif algorithm == 'kundu_v3':
selvars = [
'WTS', 'tsoc_B', 'PSC', 'Z_maps', 'F_R2_maps', 'F_S0_maps',
'Z_clmaps', 'F_R2_clmaps', 'F_S0_clmaps', 'Br_R2_clmaps',
'Br_S0_clmaps'
]
elif algorithm is None:
selvars = []
else:
raise ValueError(
'Algorithm "{0}" not recognized.'.format(algorithm))
seldict = {}
for vv in selvars:
seldict[vv] = eval(vv)
else:
seldict = None
return comptable, seldict, betas, mmix_new
def tedpca(data_cat,
data_oc,
combmode,
mask,
adaptive_mask,
t2sG,
ref_img,
tes,
algorithm='mdl',
kdaw=10.,
rdaw=1.,
out_dir='.',
verbose=False,
low_mem=False):
"""
Use principal components analysis (PCA) to identify and remove thermal
noise from multi-echo data.
Parameters
----------
data_cat : (S x E x T) array_like
Input functional data
data_oc : (S x T) array_like
Optimally combined time series data
combmode : {'t2s', 'paid'} str
How optimal combination of echos should be made, where 't2s' indicates
using the method of Posse 1999 and 'paid' indicates using the method of
Poser 2006
mask : (S,) array_like
Boolean mask array
adaptive_mask : (S,) array_like
Array where each value indicates the number of echoes with good signal
for that voxel. This mask may be thresholded; for example, with values
less than 3 set to 0.
For more information on thresholding, see `make_adaptive_mask`.
t2sG : (S,) array_like
Map of voxel-wise T2* estimates.
ref_img : :obj:`str` or img_like
Reference image to dictate how outputs are saved to disk
tes : :obj:`list`
List of echo times associated with `data_cat`, in milliseconds
algorithm : {'kundu', 'kundu-stabilize', 'mdl', 'aic', 'kic', float}, optional
Method with which to select components in TEDPCA. PCA
decomposition with the mdl, kic and aic options are based on a Moving Average
(stationary Gaussian) process and are ordered from most to least aggressive
(see Li et al., 2007).
If a float is provided, then it is assumed to represent percentage of variance
explained (0-1) to retain from PCA.
Default is 'mdl'.
kdaw : :obj:`float`, optional
Dimensionality augmentation weight for Kappa calculations. Must be a
non-negative float, or -1 (a special value). Default is 10.
rdaw : :obj:`float`, optional
Dimensionality augmentation weight for Rho calculations. Must be a
non-negative float, or -1 (a special value). Default is 1.
out_dir : :obj:`str`, optional
Output directory.
verbose : :obj:`bool`, optional
Whether to output files from fitmodels_direct or not. Default: False
low_mem : :obj:`bool`, optional
Whether to use incremental PCA (for low-memory systems) or not.
This is only compatible with the "kundu" or "kundu-stabilize" algorithms.
Default: False
Returns
-------
kept_data : (S x T) :obj:`numpy.ndarray`
Dimensionally reduced optimally combined functional data
n_components : :obj:`int`
Number of components retained from PCA decomposition
Notes
-----
====================== =================================================
Notation Meaning
====================== =================================================
:math:`\\kappa` Component pseudo-F statistic for TE-dependent
(BOLD) model.
:math:`\\rho` Component pseudo-F statistic for TE-independent
(artifact) model.
:math:`v` Voxel
:math:`V` Total number of voxels in mask
:math:`\\zeta` Something
:math:`c` Component
:math:`p` Something else
====================== =================================================
Steps:
1. Variance normalize either multi-echo or optimally combined data,
depending on settings.
2. Decompose normalized data using PCA or SVD.
3. Compute :math:`{\\kappa}` and :math:`{\\rho}`:
.. math::
{\\kappa}_c = \\frac{\\sum_{v}^V {\\zeta}_{c,v}^p * \
F_{c,v,R_2^*}}{\\sum {\\zeta}_{c,v}^p}
{\\rho}_c = \\frac{\\sum_{v}^V {\\zeta}_{c,v}^p * \
F_{c,v,S_0}}{\\sum {\\zeta}_{c,v}^p}
4. Some other stuff. Something about elbows.
5. Classify components as thermal noise if they meet both of the
following criteria:
- Nonsignificant :math:`{\\kappa}` and :math:`{\\rho}`.
- Nonsignificant variance explained.
Outputs:
This function writes out several files:
====================== =================================================
Filename Content
====================== =================================================
pca_decomposition.json PCA component table.
pca_mixing.tsv PCA mixing matrix.
pca_components.nii.gz Component weight maps.
====================== =================================================
See Also
--------
:func:`tedana.utils.make_adaptive_mask` : The function used to create the ``adaptive_mask``
parameter.
"""
if algorithm == 'kundu':
alg_str = ("followed by the Kundu component selection decision "
"tree (Kundu et al., 2013)")
RefLGR.info("Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192.")
elif algorithm == 'kundu-stabilize':
alg_str = ("followed by the 'stabilized' Kundu component "
"selection decision tree (Kundu et al., 2013)")
RefLGR.info("Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192.")
elif isinstance(algorithm, Number):
alg_str = (
"in which the number of components was determined based on a "
"variance explained threshold")
else:
alg_str = (
"based on the PCA component estimation with a Moving Average"
"(stationary Gaussian) process (Li et al., 2007)")
RefLGR.info("Li, Y.O., Adalı, T. and Calhoun, V.D., (2007). "
"Estimating the number of independent components for "
"functional magnetic resonance imaging data. "
"Human brain mapping, 28(11), pp.1251-1266.")
RepLGR.info("Principal component analysis {0} was applied to "
"the optimally combined data for dimensionality "
"reduction.".format(alg_str))
n_samp, n_echos, n_vols = data_cat.shape
LGR.info('Computing PCA of optimally combined multi-echo data')
data = data_oc[mask, :]
data_z = ((data.T - data.T.mean(axis=0)) /
data.T.std(axis=0)).T # var normalize ts
data_z = (data_z -
data_z.mean()) / data_z.std() # var normalize everything
if algorithm in ['mdl', 'aic', 'kic']:
data_img = io.new_nii_like(ref_img, utils.unmask(data, mask))
mask_img = io.new_nii_like(ref_img, mask.astype(int))
voxel_comp_weights, varex, varex_norm, comp_ts = ma_pca.ma_pca(
data_img, mask_img, algorithm)
elif isinstance(algorithm, Number):
ppca = PCA(copy=False, n_components=algorithm, svd_solver="full")
ppca.fit(data_z)
comp_ts = ppca.components_.T
varex = ppca.explained_variance_
voxel_comp_weights = np.dot(np.dot(data_z, comp_ts),
np.diag(1. / varex))
varex_norm = varex / varex.sum()
elif low_mem:
voxel_comp_weights, varex, comp_ts = low_mem_pca(data_z)
varex_norm = varex / varex.sum()
else:
ppca = PCA(copy=False, n_components=(n_vols - 1))
ppca.fit(data_z)
comp_ts = ppca.components_.T
varex = ppca.explained_variance_
voxel_comp_weights = np.dot(np.dot(data_z, comp_ts),
np.diag(1. / varex))
varex_norm = varex / varex.sum()
# Compute Kappa and Rho for PCA comps
# Normalize each component's time series
vTmixN = stats.zscore(comp_ts, axis=0)
comptable, _, _, _ = metrics.dependence_metrics(data_cat,
data_oc,
comp_ts,
adaptive_mask,
tes,
ref_img,
reindex=False,
mmixN=vTmixN,
algorithm=None,
label='mepca_',
out_dir=out_dir,
verbose=verbose)
# varex_norm from PCA overrides varex_norm from dependence_metrics,
# but we retain the original
comptable['estimated normalized variance explained'] = \
comptable['normalized variance explained']
comptable['normalized variance explained'] = varex_norm
# write component maps to 4D image
comp_ts_z = stats.zscore(comp_ts, axis=0)
comp_maps = utils.unmask(computefeats2(data_oc, comp_ts_z, mask), mask)
io.filewrite(comp_maps, op.join(out_dir, 'pca_components.nii.gz'), ref_img)
# Select components using decision tree
if algorithm == 'kundu':
comptable = kundu_tedpca(comptable,
n_echos,
kdaw,
rdaw,
stabilize=False)
elif algorithm == 'kundu-stabilize':
comptable = kundu_tedpca(comptable,
n_echos,
kdaw,
rdaw,
stabilize=True)
else:
alg_str = "variance explained-based" if isinstance(
algorithm, Number) else algorithm
LGR.info('Selected {0} components with {1} dimensionality '
'detection'.format(comptable.shape[0], alg_str))
comptable['classification'] = 'accepted'
comptable['rationale'] = ''
# Save decomposition
comp_names = [
io.add_decomp_prefix(comp,
prefix='pca',
max_value=comptable.index.max())
for comp in comptable.index.values
]
mixing_df = pd.DataFrame(data=comp_ts, columns=comp_names)
mixing_df.to_csv(op.join(out_dir, 'pca_mixing.tsv'), sep='\t', index=False)
comptable['Description'] = 'PCA fit to optimally combined data.'
mmix_dict = {}
mmix_dict['Method'] = ('Principal components analysis implemented by '
'sklearn. Components are sorted by variance '
'explained in descending order. '
'Component signs are flipped to best match the '
'data.')
io.save_comptable(comptable,
op.join(out_dir, 'pca_decomposition.json'),
label='pca',
metadata=mmix_dict)
acc = comptable[comptable.classification == 'accepted'].index.values
n_components = acc.size
voxel_kept_comp_weighted = (voxel_comp_weights[:, acc] * varex[None, acc])
kept_data = np.dot(voxel_kept_comp_weighted, comp_ts[:, acc].T)
kept_data = stats.zscore(kept_data,
axis=1) # variance normalize time series
kept_data = stats.zscore(kept_data,
axis=None) # variance normalize everything
return kept_data, n_components
def carpet_plot(optcom_ts,
denoised_ts,
hikts,
lowkts,
mask,
io_generator,
gscontrol=None):
"""Generate a set of carpet plots for the combined and denoised data.
Parameters
----------
optcom_ts, denoised_ts, hikts, lowkts : (S x T) array_like
Different types of data to plot.
mask : (S,) array-like
Binary mask used to apply to the data.
io_generator : :obj:`tedana.io.OutputGenerator`
The output generator for this workflow
gscontrol : {None, 'mir', 'gsr'} or :obj:`list`, optional
Additional denoising steps applied in the workflow.
If any gscontrol methods were applied, then additional carpet plots will be generated for
pertinent outputs from those steps.
Default is None.
"""
mask_img = io.new_nii_like(io_generator.reference_img, mask.astype(int))
optcom_img = io.new_nii_like(io_generator.reference_img, optcom_ts)
dn_img = io.new_nii_like(io_generator.reference_img, denoised_ts)
hik_img = io.new_nii_like(io_generator.reference_img, hikts)
lowk_img = io.new_nii_like(io_generator.reference_img, lowkts)
# Carpet plots
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
optcom_img,
mask_img,
figure=fig,
axes=ax,
title="Optimally Combined Data",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures", "carpet_optcom.svg"))
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
dn_img,
mask_img,
figure=fig,
axes=ax,
title="Denoised Data",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures", "carpet_denoised.svg"))
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
hik_img,
mask_img,
figure=fig,
axes=ax,
title="High-Kappa Data",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures", "carpet_accepted.svg"))
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
lowk_img,
mask_img,
figure=fig,
axes=ax,
title="Low-Kappa Data",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures", "carpet_rejected.svg"))
if (gscontrol is not None) and ("gsr" in gscontrol):
optcom_with_gs_img = io_generator.get_name("has gs combined img")
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
optcom_with_gs_img,
mask_img,
figure=fig,
axes=ax,
title="Optimally Combined Data (Pre-GSR)",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures",
"carpet_optcom_nogsr.svg"))
if (gscontrol is not None) and ("mir" in gscontrol):
mir_denoised_img = io_generator.get_name("mir denoised img")
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
mir_denoised_img,
mask_img,
figure=fig,
axes=ax,
title="Denoised Data (Post-MIR)",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures",
"carpet_denoised_mir.svg"))
mir_denoised_img = io_generator.get_name(
"ICA accepted mir denoised img")
fig, ax = plt.subplots(figsize=(14, 7))
plotting.plot_carpet(
mir_denoised_img,
mask_img,
figure=fig,
axes=ax,
title="High-Kappa Data (Post-MIR)",
)
fig.tight_layout()
fig.savefig(
os.path.join(io_generator.out_dir, "figures",
"carpet_accepted_mir.svg"))
def tedana_workflow(data, tes, mask=None, mixm=None, ctab=None, manacc=None,
tedort=False, gscontrol=None, tedpca='mle',
source_tes=-1, combmode='t2s', verbose=False, stabilize=False,
out_dir='.', fixed_seed=42, maxit=500, maxrestart=10,
debug=False, quiet=False, png=False, png_cmap='coolwarm'):
"""
Run the "canonical" TE-Dependent ANAlysis workflow.
Parameters
----------
data : :obj:`str` or :obj:`list` of :obj:`str`
Either a single z-concatenated file (single-entry list or str) or a
list of echo-specific files, in ascending order.
tes : :obj:`list`
List of echo times associated with data in milliseconds.
mask : :obj:`str`, optional
Binary mask of voxels to include in TE Dependent ANAlysis. Must be
spatially aligned with `data`. If an explicit mask is not provided,
then Nilearn's compute_epi_mask function will be used to derive a mask
from the first echo's data.
mixm : :obj:`str`, optional
File containing mixing matrix. If not provided, ME-PCA and ME-ICA are
done.
ctab : :obj:`str`, optional
File containing component table from which to extract pre-computed
classifications.
manacc : :obj:`list`, :obj:`str`, or None, optional
List of manually accepted components. Can be a list of the components,
a comma-separated string with component numbers, or None. Default is
None.
tedort : :obj:`bool`, optional
Orthogonalize rejected components w.r.t. accepted ones prior to
denoising. Default is False.
gscontrol : {None, 't1c', 'gsr'} or :obj:`list`, optional
Perform additional denoising to remove spatially diffuse noise. Default
is None.
tedpca : {'mle', 'kundu', 'kundu-stabilize'}, optional
Method with which to select components in TEDPCA. Default is 'mle'.
source_tes : :obj:`int`, optional
Source TEs for models. 0 for all, -1 for optimal combination.
Default is -1.
combmode : {'t2s'}, optional
Combination scheme for TEs: 't2s' (Posse 1999, default).
verbose : :obj:`bool`, optional
Generate intermediate and additional files. Default is False.
png : obj:'bool', optional
Generate simple plots and figures. Default is false.
png_cmap : obj:'str', optional
Name of a matplotlib colormap to be used when generating figures.
--png must still be used to request figures. Default is 'coolwarm'
out_dir : :obj:`str`, optional
Output directory.
Other Parameters
----------------
fixed_seed : :obj:`int`, optional
Value passed to ``mdp.numx_rand.seed()``.
Set to a positive integer value for reproducible ICA results;
otherwise, set to -1 for varying results across calls.
maxit : :obj:`int`, optional
Maximum number of iterations for ICA. Default is 500.
maxrestart : :obj:`int`, optional
Maximum number of attempts for ICA. If ICA fails to converge, the
fixed seed will be updated and ICA will be run again. If convergence
is achieved before maxrestart attempts, ICA will finish early.
Default is 10.
debug : :obj:`bool`, optional
Whether to run in debugging mode or not. Default is False.
quiet : :obj:`bool`, optional
If True, suppresses logging/printing of messages. Default is False.
Notes
-----
This workflow writes out several files. For a complete list of the files
generated by this workflow, please visit
https://tedana.readthedocs.io/en/latest/outputs.html
"""
out_dir = op.abspath(out_dir)
if not op.isdir(out_dir):
os.mkdir(out_dir)
if debug and not quiet:
# ensure old logs aren't over-written
basename = 'tedana_run'
extension = 'txt'
logname = op.join(out_dir, (basename + '.' + extension))
logex = op.join(out_dir, (basename + '*'))
previouslogs = glob.glob(logex)
previouslogs.sort(reverse=True)
for f in previouslogs:
previousparts = op.splitext(f)
newname = previousparts[0] + '_old' + previousparts[1]
os.rename(f, newname)
# set logging format
formatter = logging.Formatter(
'%(asctime)s\t%(name)-12s\t%(levelname)-8s\t%(message)s',
datefmt='%Y-%m-%dT%H:%M:%S')
# set up logging file and open it for writing
fh = logging.FileHandler(logname)
fh.setFormatter(formatter)
logging.basicConfig(level=logging.DEBUG,
handlers=[fh, logging.StreamHandler()])
elif quiet:
logging.basicConfig(level=logging.WARNING)
else:
logging.basicConfig(level=logging.INFO)
LGR.info('Using output directory: {}'.format(out_dir))
# ensure tes are in appropriate format
tes = [float(te) for te in tes]
n_echos = len(tes)
# Coerce gscontrol to list
if not isinstance(gscontrol, list):
gscontrol = [gscontrol]
# coerce data to samples x echos x time array
if isinstance(data, str):
data = [data]
LGR.info('Loading input data: {}'.format([f for f in data]))
catd, ref_img = io.load_data(data, n_echos=n_echos)
n_samp, n_echos, n_vols = catd.shape
LGR.debug('Resulting data shape: {}'.format(catd.shape))
if mixm is not None and op.isfile(mixm):
mixm = op.abspath(mixm)
# Allow users to re-run on same folder
if mixm != op.join(out_dir, 'meica_mix.1D'):
shutil.copyfile(mixm, op.join(out_dir, 'meica_mix.1D'))
shutil.copyfile(mixm, op.join(out_dir, op.basename(mixm)))
elif mixm is not None:
raise IOError('Argument "mixm" must be an existing file.')
if ctab is not None and op.isfile(ctab):
ctab = op.abspath(ctab)
# Allow users to re-run on same folder
if ctab != op.join(out_dir, 'comp_table_ica.txt'):
shutil.copyfile(ctab, op.join(out_dir, 'comp_table_ica.txt'))
shutil.copyfile(ctab, op.join(out_dir, op.basename(ctab)))
elif ctab is not None:
raise IOError('Argument "ctab" must be an existing file.')
if isinstance(manacc, str):
manacc = [int(comp) for comp in manacc.split(',')]
if ctab and not mixm:
LGR.warning('Argument "ctab" requires argument "mixm".')
ctab = None
elif ctab and (manacc is None):
LGR.warning('Argument "ctab" requires argument "manacc".')
ctab = None
elif manacc is not None and not mixm:
LGR.warning('Argument "manacc" requires argument "mixm".')
manacc = None
if mask is None:
LGR.info('Computing EPI mask from first echo')
first_echo_img = io.new_nii_like(ref_img, catd[:, 0, :])
mask = compute_epi_mask(first_echo_img)
else:
# TODO: add affine check
LGR.info('Using user-defined mask')
mask, masksum = utils.make_adaptive_mask(catd, mask=mask, getsum=True)
LGR.debug('Retaining {}/{} samples'.format(mask.sum(), n_samp))
if verbose:
io.filewrite(masksum, op.join(out_dir, 'adaptive_mask.nii'), ref_img)
os.chdir(out_dir)
LGR.info('Computing T2* map')
t2s, s0, t2ss, s0s, t2sG, s0G = decay.fit_decay(catd, tes, mask, masksum)
# set a hard cap for the T2* map
# anything that is 10x higher than the 99.5 %ile will be reset to 99.5 %ile
cap_t2s = stats.scoreatpercentile(t2s.flatten(), 99.5,
interpolation_method='lower')
LGR.debug('Setting cap on T2* map at {:.5f}'.format(cap_t2s * 10))
t2s[t2s > cap_t2s * 10] = cap_t2s
io.filewrite(t2s, op.join(out_dir, 't2sv.nii'), ref_img)
io.filewrite(s0, op.join(out_dir, 's0v.nii'), ref_img)
if verbose:
io.filewrite(t2ss, op.join(out_dir, 't2ss.nii'), ref_img)
io.filewrite(s0s, op.join(out_dir, 's0vs.nii'), ref_img)
io.filewrite(t2sG, op.join(out_dir, 't2svG.nii'), ref_img)
io.filewrite(s0G, op.join(out_dir, 's0vG.nii'), ref_img)
# optimally combine data
data_oc = combine.make_optcom(catd, tes, mask, t2s=t2sG, combmode=combmode)
# regress out global signal unless explicitly not desired
if 'gsr' in gscontrol:
catd, data_oc = gsc.gscontrol_raw(catd, data_oc, n_echos, ref_img)
if mixm is None:
# Identify and remove thermal noise from data
dd, n_components = decomposition.tedpca(catd, data_oc, combmode, mask,
t2s, t2sG, ref_img,
tes=tes, algorithm=tedpca,
source_tes=source_tes,
kdaw=10., rdaw=1.,
out_dir=out_dir, verbose=verbose)
mmix_orig = decomposition.tedica(dd, n_components, fixed_seed,
maxit, maxrestart)
if verbose:
np.savetxt(op.join(out_dir, '__meica_mix.1D'), mmix_orig)
if source_tes == -1:
io.filewrite(utils.unmask(dd, mask),
op.join(out_dir, 'ts_OC_whitened.nii'), ref_img)
LGR.info('Making second component selection guess from ICA results')
# Estimate betas and compute selection metrics for mixing matrix
# generated from dimensionally reduced data using full data (i.e., data
# with thermal noise)
comptable, metric_maps, betas, mmix = model.dependence_metrics(
catd, data_oc, mmix_orig, mask, t2s, tes,
ref_img, reindex=True, label='meica_', out_dir=out_dir,
algorithm='kundu_v2', verbose=verbose)
np.savetxt(op.join(out_dir, 'meica_mix.1D'), mmix)
comptable = model.kundu_metrics(comptable, metric_maps)
comptable = selection.kundu_selection_v2(comptable, n_echos, n_vols)
else:
LGR.info('Using supplied mixing matrix from ICA')
mmix_orig = np.loadtxt(op.join(out_dir, 'meica_mix.1D'))
comptable, metric_maps, betas, mmix = model.dependence_metrics(
catd, data_oc, mmix_orig, mask, t2s, tes,
ref_img, label='meica_', out_dir=out_dir,
algorithm='kundu_v2', verbose=verbose)
if ctab is None:
comptable = model.kundu_metrics(comptable, metric_maps)
comptable = selection.kundu_selection_v2(comptable, n_echos, n_vols)
else:
comptable = pd.read_csv(ctab, sep='\t', index_col='component')
comptable = selection.manual_selection(comptable, acc=manacc)
comptable.to_csv(op.join(out_dir, 'comp_table_ica.txt'), sep='\t',
index=True, index_label='component', float_format='%.6f')
if comptable[comptable.classification == 'accepted'].shape[0] == 0:
LGR.warning('No BOLD components detected! Please check data and '
'results!')
mmix_orig = mmix.copy()
if tedort:
acc_idx = comptable.loc[
~comptable.classification.str.contains('rejected')].index.values
rej_idx = comptable.loc[
comptable.classification.str.contains('rejected')].index.values
acc_ts = mmix[:, acc_idx]
rej_ts = mmix[:, rej_idx]
betas = np.linalg.lstsq(acc_ts, rej_ts, rcond=None)[0]
pred_rej_ts = np.dot(acc_ts, betas)
resid = rej_ts - pred_rej_ts
mmix[:, rej_idx] = resid
np.savetxt(op.join(out_dir, 'meica_mix_orth.1D'), mmix)
io.writeresults(data_oc, mask=mask, comptable=comptable, mmix=mmix,
n_vols=n_vols, ref_img=ref_img)
if 't1c' in gscontrol:
LGR.info('Performing T1c global signal regression to remove spatially '
'diffuse noise')
gsc.gscontrol_mmix(data_oc, mmix, mask, comptable, ref_img)
if verbose:
io.writeresults_echoes(catd, mmix, mask, comptable, ref_img)
if png:
LGR.info('Making figures folder with static component maps and '
'timecourse plots.')
# make figure folder first
if not op.isdir(op.join(out_dir, 'figures')):
os.mkdir(op.join(out_dir, 'figures'))
viz.write_comp_figs(data_oc, mask=mask, comptable=comptable,
mmix=mmix_orig, ref_img=ref_img,
out_dir=op.join(out_dir, 'figures'),
png_cmap=png_cmap)
LGR.info('Making Kappa vs Rho scatter plot')
viz.write_kappa_scatter(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Making overall summary figure')
viz.write_summary_fig(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Workflow completed')
for handler in logging.root.handlers[:]:
logging.root.removeHandler(handler)
def test_new_nii_like():
data, ref = me.load_data(fnames, n_echos=len(tes))
nimg = me.new_nii_like(ref, data)
assert isinstance(nimg, nib.Nifti1Image)
assert nimg.shape == (39, 50, 33, 3, 5)
def tedana_workflow(data, tes, mask=None, mixm=None, ctab=None, manacc=None,
tedort=False, gscontrol=None, tedpca='mle',
source_tes=-1, combmode='t2s', verbose=False, stabilize=False,
out_dir='.', fixed_seed=42, maxit=500, maxrestart=10,
debug=False, quiet=False, png=False, png_cmap='coolwarm',
low_mem=False):
"""
Run the "canonical" TE-Dependent ANAlysis workflow.
Parameters
----------
data : :obj:`str` or :obj:`list` of :obj:`str`
Either a single z-concatenated file (single-entry list or str) or a
list of echo-specific files, in ascending order.
tes : :obj:`list`
List of echo times associated with data in milliseconds.
mask : :obj:`str`, optional
Binary mask of voxels to include in TE Dependent ANAlysis. Must be
spatially aligned with `data`. If an explicit mask is not provided,
then Nilearn's compute_epi_mask function will be used to derive a mask
from the first echo's data.
mixm : :obj:`str`, optional
File containing mixing matrix. If not provided, ME-PCA and ME-ICA are
done.
ctab : :obj:`str`, optional
File containing component table from which to extract pre-computed
classifications.
manacc : :obj:`list`, :obj:`str`, or None, optional
List of manually accepted components. Can be a list of the components,
a comma-separated string with component numbers, or None. Default is
None.
tedort : :obj:`bool`, optional
Orthogonalize rejected components w.r.t. accepted ones prior to
denoising. Default is False.
gscontrol : {None, 't1c', 'gsr'} or :obj:`list`, optional
Perform additional denoising to remove spatially diffuse noise. Default
is None.
tedpca : {'mle', 'kundu', 'kundu-stabilize'}, optional
Method with which to select components in TEDPCA. Default is 'mle'.
source_tes : :obj:`int`, optional
Source TEs for models. 0 for all, -1 for optimal combination.
Default is -1.
combmode : {'t2s'}, optional
Combination scheme for TEs: 't2s' (Posse 1999, default).
verbose : :obj:`bool`, optional
Generate intermediate and additional files. Default is False.
png : obj:'bool', optional
Generate simple plots and figures. Default is false.
png_cmap : obj:'str', optional
Name of a matplotlib colormap to be used when generating figures.
--png must still be used to request figures. Default is 'coolwarm'
out_dir : :obj:`str`, optional
Output directory.
Other Parameters
----------------
fixed_seed : :obj:`int`, optional
Value passed to ``mdp.numx_rand.seed()``.
Set to a positive integer value for reproducible ICA results;
otherwise, set to -1 for varying results across calls.
maxit : :obj:`int`, optional
Maximum number of iterations for ICA. Default is 500.
maxrestart : :obj:`int`, optional
Maximum number of attempts for ICA. If ICA fails to converge, the
fixed seed will be updated and ICA will be run again. If convergence
is achieved before maxrestart attempts, ICA will finish early.
Default is 10.
low_mem : :obj:`bool`, optional
Enables low-memory processing, including the use of IncrementalPCA.
May increase workflow duration. Default is False.
debug : :obj:`bool`, optional
Whether to run in debugging mode or not. Default is False.
quiet : :obj:`bool`, optional
If True, suppresses logging/printing of messages. Default is False.
Notes
-----
This workflow writes out several files. For a complete list of the files
generated by this workflow, please visit
https://tedana.readthedocs.io/en/latest/outputs.html
"""
out_dir = op.abspath(out_dir)
if not op.isdir(out_dir):
os.mkdir(out_dir)
# boilerplate
refs = []
basename = 'report'
extension = 'txt'
repname = op.join(out_dir, (basename + '.' + extension))
repex = op.join(out_dir, (basename + '*'))
previousreps = glob(repex)
previousreps.sort(reverse=True)
for f in previousreps:
previousparts = op.splitext(f)
newname = previousparts[0] + '_old' + previousparts[1]
os.rename(f, newname)
if debug and not quiet:
# ensure old logs aren't over-written
basename = 'tedana_run'
extension = 'txt'
logname = op.join(out_dir, (basename + '.' + extension))
logex = op.join(out_dir, (basename + '*'))
previouslogs = glob(logex)
previouslogs.sort(reverse=True)
for f in previouslogs:
previousparts = op.splitext(f)
newname = previousparts[0] + '_old' + previousparts[1]
os.rename(f, newname)
# set logging format
formatter = logging.Formatter(
'%(asctime)s\t%(name)-12s\t%(levelname)-8s\t%(message)s',
datefmt='%Y-%m-%dT%H:%M:%S')
# set up logging file and open it for writing
fh = logging.FileHandler(logname)
fh.setFormatter(formatter)
logging.basicConfig(level=logging.DEBUG,
handlers=[fh, logging.StreamHandler()])
elif quiet:
logging.basicConfig(level=logging.WARNING)
else:
logging.basicConfig(level=logging.INFO)
LGR.info('Using output directory: {}'.format(out_dir))
# ensure tes are in appropriate format
tes = [float(te) for te in tes]
n_echos = len(tes)
# Coerce gscontrol to list
if not isinstance(gscontrol, list):
gscontrol = [gscontrol]
# coerce data to samples x echos x time array
if isinstance(data, str):
data = [data]
LGR.info('Loading input data: {}'.format([f for f in data]))
catd, ref_img = io.load_data(data, n_echos=n_echos)
n_samp, n_echos, n_vols = catd.shape
LGR.debug('Resulting data shape: {}'.format(catd.shape))
# check if TR is 0
img_t_r = ref_img.header.get_zooms()[-1]
if img_t_r == 0 and png:
raise IOError('Dataset has a TR of 0. This indicates incorrect'
' header information. To correct this, we recommend'
' using this snippet:'
'\n'
'https://gist.github.com/jbteves/032c87aeb080dd8de8861cb151bff5d6'
'\n'
'to correct your TR to the value it should be.')
if mixm is not None and op.isfile(mixm):
mixm = op.abspath(mixm)
# Allow users to re-run on same folder
if mixm != op.join(out_dir, 'meica_mix.1D'):
shutil.copyfile(mixm, op.join(out_dir, 'meica_mix.1D'))
shutil.copyfile(mixm, op.join(out_dir, op.basename(mixm)))
elif mixm is not None:
raise IOError('Argument "mixm" must be an existing file.')
if ctab is not None and op.isfile(ctab):
ctab = op.abspath(ctab)
# Allow users to re-run on same folder
if ctab != op.join(out_dir, 'comp_table_ica.txt'):
shutil.copyfile(ctab, op.join(out_dir, 'comp_table_ica.txt'))
shutil.copyfile(ctab, op.join(out_dir, op.basename(ctab)))
elif ctab is not None:
raise IOError('Argument "ctab" must be an existing file.')
if isinstance(manacc, str):
manacc = [int(comp) for comp in manacc.split(',')]
if ctab and not mixm:
LGR.warning('Argument "ctab" requires argument "mixm".')
ctab = None
elif ctab and (manacc is None):
LGR.warning('Argument "ctab" requires argument "manacc".')
ctab = None
elif manacc is not None and not mixm:
LGR.warning('Argument "manacc" requires argument "mixm".')
manacc = None
bp_str = ("TE-dependence analysis was performed on input data.")
if mask is None:
LGR.info('Computing EPI mask from first echo')
first_echo_img = io.new_nii_like(ref_img, catd[:, 0, :])
mask = compute_epi_mask(first_echo_img)
bp_str += (" An initial mask was generated from the first echo using "
"nilearn's compute_epi_mask function.")
else:
# TODO: add affine check
LGR.info('Using user-defined mask')
bp_str += (" A user-defined mask was applied to the data.")
mask, masksum = utils.make_adaptive_mask(catd, mask=mask, getsum=True)
bp_str += (" An adaptive mask was then generated, in which each voxel's "
"value reflects the number of echoes with 'good' data.")
LGR.debug('Retaining {}/{} samples'.format(mask.sum(), n_samp))
if verbose:
io.filewrite(masksum, op.join(out_dir, 'adaptive_mask.nii'), ref_img)
os.chdir(out_dir)
LGR.info('Computing T2* map')
t2s, s0, t2ss, s0s, t2sG, s0G = decay.fit_decay(catd, tes, mask, masksum)
bp_str += (" A monoexponential model was fit to the data at each voxel "
"using log-linear regression in order to estimate T2* and S0 "
"maps. For each voxel, the value from the adaptive mask was "
"used to determine which echoes would be used to estimate T2* "
"and S0.")
# set a hard cap for the T2* map
# anything that is 10x higher than the 99.5 %ile will be reset to 99.5 %ile
cap_t2s = stats.scoreatpercentile(t2s.flatten(), 99.5,
interpolation_method='lower')
LGR.debug('Setting cap on T2* map at {:.5f}'.format(cap_t2s * 10))
t2s[t2s > cap_t2s * 10] = cap_t2s
io.filewrite(t2s, op.join(out_dir, 't2sv.nii'), ref_img)
io.filewrite(s0, op.join(out_dir, 's0v.nii'), ref_img)
if verbose:
io.filewrite(t2ss, op.join(out_dir, 't2ss.nii'), ref_img)
io.filewrite(s0s, op.join(out_dir, 's0vs.nii'), ref_img)
io.filewrite(t2sG, op.join(out_dir, 't2svG.nii'), ref_img)
io.filewrite(s0G, op.join(out_dir, 's0vG.nii'), ref_img)
# optimally combine data
data_oc = combine.make_optcom(catd, tes, mask, t2s=t2sG, combmode=combmode)
if combmode == 't2s':
cm_str = "'t2s' (Posse et al., 1999)"
refs += ["Posse, S., Wiese, S., Gembris, D., Mathiak, K., Kessler, "
"C., Grosse‐Ruyken, M. L., ... & Kiselev, V. G. (1999). "
"Enhancement of BOLD‐contrast sensitivity by single‐shot "
"multi‐echo functional MR imaging. Magnetic Resonance in "
"Medicine: An Official Journal of the International Society "
"for Magnetic Resonance in Medicine, 42(1), 87-97."]
bp_str += (" Multi-echo data were then optimally combined using the {0} "
"combination method.").format(cm_str)
# regress out global signal unless explicitly not desired
if 'gsr' in gscontrol:
catd, data_oc = gsc.gscontrol_raw(catd, data_oc, n_echos, ref_img)
bp_str += (" Global signal regression was applied to the multi-echo "
"and optimally combined datasets.")
if mixm is None:
# Identify and remove thermal noise from data
dd, n_components = decomposition.tedpca(catd, data_oc, combmode, mask,
t2s, t2sG, ref_img,
tes=tes, algorithm=tedpca,
source_tes=source_tes,
kdaw=10., rdaw=1.,
out_dir=out_dir,
verbose=verbose,
low_mem=low_mem)
if tedpca == 'mle':
alg_str = "using MLE dimensionality estimation (Minka, 2001)"
refs += ["Minka, T. P. (2001). Automatic choice of dimensionality "
"for PCA. In Advances in neural information processing "
"systems (pp. 598-604)."]
elif tedpca == 'kundu':
alg_str = ("followed by the Kundu component selection decision "
"tree (Kundu et al., 2013)")
refs += ["Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192."]
elif tedpca == 'kundu-stabilize':
alg_str = ("followed by the 'stabilized' Kundu component "
"selection decision tree (Kundu et al., 2013)")
refs += ["Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192."]
if source_tes == -1:
dat_str = "the optimally combined data"
elif source_tes == 0:
dat_str = "the z-concatenated multi-echo data"
else:
dat_str = "a z-concatenated subset of echoes from the input data"
bp_str += (" Principal component analysis {0} was applied to "
"{1} for dimensionality reduction.").format(alg_str,
dat_str)
mmix_orig = decomposition.tedica(dd, n_components, fixed_seed,
maxit, maxrestart)
bp_str += (" Independent component analysis was then used to "
"decompose the dimensionally reduced dataset.")
if verbose:
np.savetxt(op.join(out_dir, '__meica_mix.1D'), mmix_orig)
if source_tes == -1:
io.filewrite(utils.unmask(dd, mask),
op.join(out_dir, 'ts_OC_whitened.nii'), ref_img)
LGR.info('Making second component selection guess from ICA results')
# Estimate betas and compute selection metrics for mixing matrix
# generated from dimensionally reduced data using full data (i.e., data
# with thermal noise)
comptable, metric_maps, betas, mmix = metrics.dependence_metrics(
catd, data_oc, mmix_orig, t2s, tes,
ref_img, reindex=True, label='meica_', out_dir=out_dir,
algorithm='kundu_v2', verbose=verbose)
bp_str += (" A series of TE-dependence metrics were calculated for "
"each ICA component, including Kappa, Rho, and variance "
"explained.")
np.savetxt(op.join(out_dir, 'meica_mix.1D'), mmix)
comptable = metrics.kundu_metrics(comptable, metric_maps)
comptable = selection.kundu_selection_v2(comptable, n_echos, n_vols)
bp_str += (" Next, component selection was performed to identify "
"BOLD (TE-dependent), non-BOLD (TE-independent), and "
"uncertain (low-variance) components using the Kundu "
"decision tree (v2.5; Kundu et al., 2013).")
refs += ["Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192."]
else:
LGR.info('Using supplied mixing matrix from ICA')
mmix_orig = np.loadtxt(op.join(out_dir, 'meica_mix.1D'))
comptable, metric_maps, betas, mmix = metrics.dependence_metrics(
catd, data_oc, mmix_orig, t2s, tes,
ref_img, label='meica_', out_dir=out_dir,
algorithm='kundu_v2', verbose=verbose)
if ctab is None:
comptable = metrics.kundu_metrics(comptable, metric_maps)
comptable = selection.kundu_selection_v2(comptable, n_echos, n_vols)
bp_str += (" Next, component selection was performed to identify "
"BOLD (TE-dependent), non-BOLD (TE-independent), and "
"uncertain (low-variance) components using the Kundu "
"decision tree (v2.5; Kundu et al., 2013).")
refs += ["Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192."]
else:
comptable = pd.read_csv(ctab, sep='\t', index_col='component')
comptable = selection.manual_selection(comptable, acc=manacc)
bp_str += (" Next, components were manually classified as "
"BOLD (TE-dependent), non-BOLD (TE-independent), or "
"uncertain (low-variance).")
comptable.to_csv(op.join(out_dir, 'comp_table_ica.txt'), sep='\t',
index=True, index_label='component', float_format='%.6f')
if comptable[comptable.classification == 'accepted'].shape[0] == 0:
LGR.warning('No BOLD components detected! Please check data and '
'results!')
mmix_orig = mmix.copy()
if tedort:
acc_idx = comptable.loc[
~comptable.classification.str.contains('rejected')].index.values
rej_idx = comptable.loc[
comptable.classification.str.contains('rejected')].index.values
acc_ts = mmix[:, acc_idx]
rej_ts = mmix[:, rej_idx]
betas = np.linalg.lstsq(acc_ts, rej_ts, rcond=None)[0]
pred_rej_ts = np.dot(acc_ts, betas)
resid = rej_ts - pred_rej_ts
mmix[:, rej_idx] = resid
np.savetxt(op.join(out_dir, 'meica_mix_orth.1D'), mmix)
bp_str += (" Rejected components' time series were then "
"orthogonalized with respect to accepted components' time "
"series.")
io.writeresults(data_oc, mask=mask, comptable=comptable, mmix=mmix,
n_vols=n_vols, ref_img=ref_img)
if 't1c' in gscontrol:
LGR.info('Performing T1c global signal regression to remove spatially '
'diffuse noise')
gsc.gscontrol_mmix(data_oc, mmix, mask, comptable, ref_img)
bp_str += (" T1c global signal regression was then applied to the "
"data in order to remove spatially diffuse noise.")
if verbose:
io.writeresults_echoes(catd, mmix, mask, comptable, ref_img)
if png:
LGR.info('Making figures folder with static component maps and '
'timecourse plots.')
# make figure folder first
if not op.isdir(op.join(out_dir, 'figures')):
os.mkdir(op.join(out_dir, 'figures'))
viz.write_comp_figs(data_oc, mask=mask, comptable=comptable,
mmix=mmix_orig, ref_img=ref_img,
out_dir=op.join(out_dir, 'figures'),
png_cmap=png_cmap)
LGR.info('Making Kappa vs Rho scatter plot')
viz.write_kappa_scatter(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Making overall summary figure')
viz.write_summary_fig(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Workflow completed')
bp_str += ("\n\nThis workflow used numpy (Van Der Walt, Colbert, & "
"Varoquaux, 2011), scipy (Jones et al., 2001), pandas "
"(McKinney, 2010), scikit-learn (Pedregosa et al., 2011), "
"nilearn, and nibabel (Brett et al., 2019).")
refs += ["Van Der Walt, S., Colbert, S. C., & Varoquaux, G. (2011). The "
"NumPy array: a structure for efficient numerical computation. "
"Computing in Science & Engineering, 13(2), 22.",
"Jones E, Oliphant E, Peterson P, et al. SciPy: Open Source "
"Scientific Tools for Python, 2001-, http://www.scipy.org/",
"McKinney, W. (2010, June). Data structures for statistical "
"computing in python. In Proceedings of the 9th Python in "
"Science Conference (Vol. 445, pp. 51-56).",
"Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., "
"Thirion, B., Grisel, O., ... & Vanderplas, J. (2011). "
"Scikit-learn: Machine learning in Python. Journal of machine "
"learning research, 12(Oct), 2825-2830.",
"Brett, M., Markiewicz, C. J., Hanke, M., Côté, M.-A., "
"Cipollini, B., McCarthy, P., … freec84. (2019, May 28). "
"nipy/nibabel. Zenodo. http://doi.org/10.5281/zenodo.3233118"]
bp_str += ("\n\nThis workflow also used the Dice similarity index "
"(Dice, 1945; Sørensen, 1948).")
refs += ["Dice, L. R. (1945). Measures of the amount of ecologic "
"association between species. Ecology, 26(3), 297-302.",
"Sørensen, T. J. (1948). A method of establishing groups of "
"equal amplitude in plant sociology based on similarity of "
"species content and its application to analyses of the "
"vegetation on Danish commons. I kommission hos E. Munksgaard."]
bp_str += '\n\nReferences\n\n'
refs = sorted(list(set(refs)))
bp_str += '\n\n'.join(refs)
with open(repname, 'w') as fo:
fo.write(bp_str)
for handler in logging.root.handlers[:]:
logging.root.removeHandler(handler)
def tedana_workflow(data,
tes,
out_dir='.',
mask=None,
fittype='loglin',
combmode='t2s',
tedpca='mdl',
fixed_seed=42,
maxit=500,
maxrestart=10,
tedort=False,
gscontrol=None,
no_png=False,
png_cmap='coolwarm',
verbose=False,
low_mem=False,
debug=False,
quiet=False,
t2smap=None,
mixm=None,
ctab=None,
manacc=None):
"""
Run the "canonical" TE-Dependent ANAlysis workflow.
Parameters
----------
data : :obj:`str` or :obj:`list` of :obj:`str`
Either a single z-concatenated file (single-entry list or str) or a
list of echo-specific files, in ascending order.
tes : :obj:`list`
List of echo times associated with data in milliseconds.
out_dir : :obj:`str`, optional
Output directory.
mask : :obj:`str` or None, optional
Binary mask of voxels to include in TE Dependent ANAlysis. Must be
spatially aligned with `data`. If an explicit mask is not provided,
then Nilearn's compute_epi_mask function will be used to derive a mask
from the first echo's data.
fittype : {'loglin', 'curvefit'}, optional
Monoexponential fitting method. 'loglin' uses the the default linear
fit to the log of the data. 'curvefit' uses a monoexponential fit to
the raw data, which is slightly slower but may be more accurate.
Default is 'loglin'.
combmode : {'t2s'}, optional
Combination scheme for TEs: 't2s' (Posse 1999, default).
tedpca : {'kundu', 'kundu-stabilize', 'mdl', 'aic', 'kic'}, optional
Method with which to select components in TEDPCA. Default is 'mdl'.
tedort : :obj:`bool`, optional
Orthogonalize rejected components w.r.t. accepted ones prior to
denoising. Default is False.
gscontrol : {None, 't1c', 'gsr'} or :obj:`list`, optional
Perform additional denoising to remove spatially diffuse noise. Default
is None.
verbose : :obj:`bool`, optional
Generate intermediate and additional files. Default is False.
no_png : obj:'bool', optional
Do not generate .png plots and figures. Default is false.
png_cmap : obj:'str', optional
Name of a matplotlib colormap to be used when generating figures.
Cannot be used with --no-png. Default is 'coolwarm'.
t2smap : :obj:`str`, optional
Precalculated T2* map in the same space as the input data.
mixm : :obj:`str` or None, optional
File containing mixing matrix, to be used when re-running the workflow.
If not provided, ME-PCA and ME-ICA are done. Default is None.
ctab : :obj:`str` or None, optional
File containing component table from which to extract pre-computed
classifications, to be used with 'mixm' when re-running the workflow.
Default is None.
manacc : :obj:`list`, :obj:`str`, or None, optional
List of manually accepted components. Can be a list of the components,
a comma-separated string with component numbers, or None. Default is
None.
Other Parameters
----------------
fixed_seed : :obj:`int`, optional
Value passed to ``mdp.numx_rand.seed()``.
Set to a positive integer value for reproducible ICA results;
otherwise, set to -1 for varying results across calls.
maxit : :obj:`int`, optional
Maximum number of iterations for ICA. Default is 500.
maxrestart : :obj:`int`, optional
Maximum number of attempts for ICA. If ICA fails to converge, the
fixed seed will be updated and ICA will be run again. If convergence
is achieved before maxrestart attempts, ICA will finish early.
Default is 10.
low_mem : :obj:`bool`, optional
Enables low-memory processing, including the use of IncrementalPCA.
May increase workflow duration. Default is False.
debug : :obj:`bool`, optional
Whether to run in debugging mode or not. Default is False.
quiet : :obj:`bool`, optional
If True, suppresses logging/printing of messages. Default is False.
Notes
-----
This workflow writes out several files. For a complete list of the files
generated by this workflow, please visit
https://tedana.readthedocs.io/en/latest/outputs.html
"""
out_dir = op.abspath(out_dir)
if not op.isdir(out_dir):
os.mkdir(out_dir)
# boilerplate
basename = 'report'
extension = 'txt'
repname = op.join(out_dir, (basename + '.' + extension))
repex = op.join(out_dir, (basename + '*'))
previousreps = glob(repex)
previousreps.sort(reverse=True)
for f in previousreps:
previousparts = op.splitext(f)
newname = previousparts[0] + '_old' + previousparts[1]
os.rename(f, newname)
refname = op.join(out_dir, '_references.txt')
# create logfile name
basename = 'tedana_'
extension = 'tsv'
start_time = datetime.datetime.now().strftime('%Y-%m-%dT%H%M%S')
logname = op.join(out_dir, (basename + start_time + '.' + extension))
# set logging format
log_formatter = logging.Formatter(
'%(asctime)s\t%(name)-12s\t%(levelname)-8s\t%(message)s',
datefmt='%Y-%m-%dT%H:%M:%S')
text_formatter = logging.Formatter('%(message)s')
# set up logging file and open it for writing
log_handler = logging.FileHandler(logname)
log_handler.setFormatter(log_formatter)
# Removing handlers after basicConfig doesn't work, so we use filters
# for the relevant handlers themselves.
log_handler.addFilter(ContextFilter())
sh = logging.StreamHandler()
sh.addFilter(ContextFilter())
if quiet:
logging.basicConfig(level=logging.WARNING, handlers=[log_handler, sh])
elif debug:
logging.basicConfig(level=logging.DEBUG, handlers=[log_handler, sh])
else:
logging.basicConfig(level=logging.INFO, handlers=[log_handler, sh])
# Loggers for report and references
rep_handler = logging.FileHandler(repname)
rep_handler.setFormatter(text_formatter)
ref_handler = logging.FileHandler(refname)
ref_handler.setFormatter(text_formatter)
RepLGR.setLevel(logging.INFO)
RepLGR.addHandler(rep_handler)
RepLGR.setLevel(logging.INFO)
RefLGR.addHandler(ref_handler)
LGR.info('Using output directory: {}'.format(out_dir))
# ensure tes are in appropriate format
tes = [float(te) for te in tes]
n_echos = len(tes)
# Coerce gscontrol to list
if not isinstance(gscontrol, list):
gscontrol = [gscontrol]
LGR.info('Loading input data: {}'.format([f for f in data]))
catd, ref_img = io.load_data(data, n_echos=n_echos)
n_samp, n_echos, n_vols = catd.shape
LGR.debug('Resulting data shape: {}'.format(catd.shape))
if no_png and (png_cmap != 'coolwarm'):
LGR.warning('Overriding --no-png since --png-cmap provided.')
no_png = False
# check if TR is 0
img_t_r = ref_img.header.get_zooms()[-1]
if img_t_r == 0 and not no_png:
raise IOError(
'Dataset has a TR of 0. This indicates incorrect'
' header information. To correct this, we recommend'
' using this snippet:'
'\n'
'https://gist.github.com/jbteves/032c87aeb080dd8de8861cb151bff5d6'
'\n'
'to correct your TR to the value it should be.')
if mixm is not None and op.isfile(mixm):
mixm = op.abspath(mixm)
# Allow users to re-run on same folder
if mixm != op.join(out_dir, 'ica_mixing.tsv'):
shutil.copyfile(mixm, op.join(out_dir, 'ica_mixing.tsv'))
shutil.copyfile(mixm, op.join(out_dir, op.basename(mixm)))
elif mixm is not None:
raise IOError('Argument "mixm" must be an existing file.')
if ctab is not None and op.isfile(ctab):
ctab = op.abspath(ctab)
# Allow users to re-run on same folder
if ctab != op.join(out_dir, 'ica_decomposition.json'):
shutil.copyfile(ctab, op.join(out_dir, 'ica_decomposition.json'))
shutil.copyfile(ctab, op.join(out_dir, op.basename(ctab)))
elif ctab is not None:
raise IOError('Argument "ctab" must be an existing file.')
if isinstance(manacc, str):
manacc = [int(comp) for comp in manacc.split(',')]
if ctab and not mixm:
LGR.warning('Argument "ctab" requires argument "mixm".')
ctab = None
elif manacc is not None and not mixm:
LGR.warning('Argument "manacc" requires argument "mixm".')
manacc = None
if t2smap is not None and op.isfile(t2smap):
t2smap = op.abspath(t2smap)
# Allow users to re-run on same folder
if t2smap != op.join(out_dir, 't2sv.nii.gz'):
shutil.copyfile(t2smap, op.join(out_dir, 't2sv.nii.gz'))
shutil.copyfile(t2smap, op.join(out_dir, op.basename(t2smap)))
elif t2smap is not None:
raise IOError('Argument "t2smap" must be an existing file.')
RepLGR.info("TE-dependence analysis was performed on input data.")
if mask and not t2smap:
# TODO: add affine check
LGR.info('Using user-defined mask')
RepLGR.info("A user-defined mask was applied to the data.")
elif t2smap and not mask:
LGR.info('Using user-defined T2* map to generate mask')
t2s_limited = utils.load_image(t2smap)
t2s_full = t2s_limited.copy()
mask = (t2s_limited != 0).astype(int)
elif t2smap and mask:
LGR.info('Combining user-defined mask and T2* map to generate mask')
t2s_limited = utils.load_image(t2smap)
t2s_full = t2s_limited.copy()
mask = utils.load_image(mask)
mask[t2s_limited == 0] = 0 # reduce mask based on T2* map
else:
LGR.info('Computing EPI mask from first echo')
first_echo_img = io.new_nii_like(ref_img, catd[:, 0, :])
mask = compute_epi_mask(first_echo_img)
RepLGR.info("An initial mask was generated from the first echo using "
"nilearn's compute_epi_mask function.")
mask, masksum = utils.make_adaptive_mask(catd, mask=mask, getsum=True)
LGR.debug('Retaining {}/{} samples'.format(mask.sum(), n_samp))
io.filewrite(masksum, op.join(out_dir, 'adaptive_mask.nii'), ref_img)
if t2smap is None:
LGR.info('Computing T2* map')
t2s_limited, s0_limited, t2s_full, s0_full = decay.fit_decay(
catd, tes, mask, masksum, fittype)
# set a hard cap for the T2* map
# anything that is 10x higher than the 99.5 %ile will be reset to 99.5 %ile
cap_t2s = stats.scoreatpercentile(t2s_limited.flatten(),
99.5,
interpolation_method='lower')
LGR.debug('Setting cap on T2* map at {:.5f}'.format(cap_t2s * 10))
t2s_limited[t2s_limited > cap_t2s * 10] = cap_t2s
io.filewrite(t2s_limited, op.join(out_dir, 't2sv.nii'), ref_img)
io.filewrite(s0_limited, op.join(out_dir, 's0v.nii'), ref_img)
if verbose:
io.filewrite(t2s_full, op.join(out_dir, 't2svG.nii'), ref_img)
io.filewrite(s0_full, op.join(out_dir, 's0vG.nii'), ref_img)
# optimally combine data
data_oc = combine.make_optcom(catd,
tes,
mask,
t2s=t2s_full,
combmode=combmode)
# regress out global signal unless explicitly not desired
if 'gsr' in gscontrol:
catd, data_oc = gsc.gscontrol_raw(catd,
data_oc,
n_echos,
ref_img,
out_dir=out_dir)
if mixm is None:
# Identify and remove thermal noise from data
dd, n_components = decomposition.tedpca(catd,
data_oc,
combmode,
mask,
t2s_limited,
t2s_full,
ref_img,
tes=tes,
algorithm=tedpca,
kdaw=10.,
rdaw=1.,
out_dir=out_dir,
verbose=verbose,
low_mem=low_mem)
mmix_orig = decomposition.tedica(dd, n_components, fixed_seed, maxit,
maxrestart)
if verbose:
io.filewrite(utils.unmask(dd, mask),
op.join(out_dir, 'ts_OC_whitened.nii.gz'), ref_img)
LGR.info('Making second component selection guess from ICA results')
# Estimate betas and compute selection metrics for mixing matrix
# generated from dimensionally reduced data using full data (i.e., data
# with thermal noise)
comptable, metric_maps, betas, mmix = metrics.dependence_metrics(
catd,
data_oc,
mmix_orig,
t2s_limited,
tes,
ref_img,
reindex=True,
label='meica_',
out_dir=out_dir,
algorithm='kundu_v2',
verbose=verbose)
comp_names = [
io.add_decomp_prefix(comp,
prefix='ica',
max_value=comptable.index.max())
for comp in comptable.index.values
]
mixing_df = pd.DataFrame(data=mmix, columns=comp_names)
mixing_df.to_csv(op.join(out_dir, 'ica_mixing.tsv'),
sep='\t',
index=False)
betas_oc = utils.unmask(computefeats2(data_oc, mmix, mask), mask)
io.filewrite(betas_oc, op.join(out_dir, 'ica_components.nii.gz'),
ref_img)
comptable = metrics.kundu_metrics(comptable, metric_maps)
comptable = selection.kundu_selection_v2(comptable, n_echos, n_vols)
else:
LGR.info('Using supplied mixing matrix from ICA')
mmix_orig = pd.read_table(op.join(out_dir, 'ica_mixing.tsv')).values
if ctab is None:
comptable, metric_maps, betas, mmix = metrics.dependence_metrics(
catd,
data_oc,
mmix_orig,
t2s_limited,
tes,
ref_img,
label='meica_',
out_dir=out_dir,
algorithm='kundu_v2',
verbose=verbose)
comptable = metrics.kundu_metrics(comptable, metric_maps)
comptable = selection.kundu_selection_v2(comptable, n_echos,
n_vols)
else:
mmix = mmix_orig.copy()
comptable = io.load_comptable(ctab)
if manacc is not None:
comptable = selection.manual_selection(comptable, acc=manacc)
betas_oc = utils.unmask(computefeats2(data_oc, mmix, mask), mask)
io.filewrite(betas_oc, op.join(out_dir, 'ica_components.nii.gz'),
ref_img)
# Save decomposition
comptable[
'Description'] = 'ICA fit to dimensionally-reduced optimally combined data.'
mmix_dict = {}
mmix_dict['Method'] = ('Independent components analysis with FastICA '
'algorithm implemented by sklearn. Components '
'are sorted by Kappa in descending order. '
'Component signs are flipped to best match the '
'data.')
io.save_comptable(comptable,
op.join(out_dir, 'ica_decomposition.json'),
label='ica',
metadata=mmix_dict)
if comptable[comptable.classification == 'accepted'].shape[0] == 0:
LGR.warning('No BOLD components detected! Please check data and '
'results!')
mmix_orig = mmix.copy()
if tedort:
acc_idx = comptable.loc[~comptable.classification.str.
contains('rejected')].index.values
rej_idx = comptable.loc[comptable.classification.str.contains(
'rejected')].index.values
acc_ts = mmix[:, acc_idx]
rej_ts = mmix[:, rej_idx]
betas = np.linalg.lstsq(acc_ts, rej_ts, rcond=None)[0]
pred_rej_ts = np.dot(acc_ts, betas)
resid = rej_ts - pred_rej_ts
mmix[:, rej_idx] = resid
comp_names = [
io.add_decomp_prefix(comp,
prefix='ica',
max_value=comptable.index.max())
for comp in comptable.index.values
]
mixing_df = pd.DataFrame(data=mmix, columns=comp_names)
mixing_df.to_csv(op.join(out_dir, 'ica_orth_mixing.tsv'),
sep='\t',
index=False)
RepLGR.info("Rejected components' time series were then "
"orthogonalized with respect to accepted components' time "
"series.")
io.writeresults(data_oc,
mask=mask,
comptable=comptable,
mmix=mmix,
n_vols=n_vols,
ref_img=ref_img,
out_dir=out_dir)
if 't1c' in gscontrol:
gsc.gscontrol_mmix(data_oc,
mmix,
mask,
comptable,
ref_img,
out_dir=out_dir)
if verbose:
io.writeresults_echoes(catd,
mmix,
mask,
comptable,
ref_img,
out_dir=out_dir)
if not no_png:
LGR.info('Making figures folder with static component maps and '
'timecourse plots.')
# make figure folder first
if not op.isdir(op.join(out_dir, 'figures')):
os.mkdir(op.join(out_dir, 'figures'))
viz.write_comp_figs(data_oc,
mask=mask,
comptable=comptable,
mmix=mmix_orig,
ref_img=ref_img,
out_dir=op.join(out_dir, 'figures'),
png_cmap=png_cmap)
LGR.info('Making Kappa vs Rho scatter plot')
viz.write_kappa_scatter(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Making Kappa/Rho scree plot')
viz.write_kappa_scree(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Making overall summary figure')
viz.write_summary_fig(comptable=comptable,
out_dir=op.join(out_dir, 'figures'))
LGR.info('Workflow completed')
RepLGR.info("This workflow used numpy (Van Der Walt, Colbert, & "
"Varoquaux, 2011), scipy (Jones et al., 2001), pandas "
"(McKinney, 2010), scikit-learn (Pedregosa et al., 2011), "
"nilearn, and nibabel (Brett et al., 2019).")
RefLGR.info(
"Van Der Walt, S., Colbert, S. C., & Varoquaux, G. (2011). The "
"NumPy array: a structure for efficient numerical computation. "
"Computing in Science & Engineering, 13(2), 22.")
RefLGR.info("Jones E, Oliphant E, Peterson P, et al. SciPy: Open Source "
"Scientific Tools for Python, 2001-, http://www.scipy.org/")
RefLGR.info("McKinney, W. (2010, June). Data structures for statistical "
"computing in python. In Proceedings of the 9th Python in "
"Science Conference (Vol. 445, pp. 51-56).")
RefLGR.info("Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., "
"Thirion, B., Grisel, O., ... & Vanderplas, J. (2011). "
"Scikit-learn: Machine learning in Python. Journal of machine "
"learning research, 12(Oct), 2825-2830.")
RefLGR.info("Brett, M., Markiewicz, C. J., Hanke, M., Côté, M.-A., "
"Cipollini, B., McCarthy, P., … freec84. (2019, May 28). "
"nipy/nibabel. Zenodo. http://doi.org/10.5281/zenodo.3233118")
RepLGR.info("This workflow also used the Dice similarity index "
"(Dice, 1945; Sørensen, 1948).")
RefLGR.info("Dice, L. R. (1945). Measures of the amount of ecologic "
"association between species. Ecology, 26(3), 297-302.")
RefLGR.info(
"Sørensen, T. J. (1948). A method of establishing groups of "
"equal amplitude in plant sociology based on similarity of "
"species content and its application to analyses of the "
"vegetation on Danish commons. I kommission hos E. Munksgaard.")
with open(repname, 'r') as fo:
report = [line.rstrip() for line in fo.readlines()]
report = ' '.join(report)
with open(refname, 'r') as fo:
reference_list = sorted(list(set(fo.readlines())))
references = '\n'.join(reference_list)
report += '\n\nReferences\n' + references
with open(repname, 'w') as fo:
fo.write(report)
os.remove(refname)
for handler in logging.root.handlers[:]:
logging.root.removeHandler(handler)
def tedana_workflow(data,
tes,
out_dir='.',
mask=None,
convention='bids',
prefix='',
fittype='loglin',
combmode='t2s',
tedpca='mdl',
fixed_seed=42,
maxit=500,
maxrestart=10,
tedort=False,
gscontrol=None,
no_reports=False,
png_cmap='coolwarm',
verbose=False,
low_mem=False,
debug=False,
quiet=False,
t2smap=None,
mixm=None,
ctab=None,
manacc=None):
"""
Run the "canonical" TE-Dependent ANAlysis workflow.
Parameters
----------
data : :obj:`str` or :obj:`list` of :obj:`str`
Either a single z-concatenated file (single-entry list or str) or a
list of echo-specific files, in ascending order.
tes : :obj:`list`
List of echo times associated with data in milliseconds.
out_dir : :obj:`str`, optional
Output directory.
mask : :obj:`str` or None, optional
Binary mask of voxels to include in TE Dependent ANAlysis. Must be
spatially aligned with `data`. If an explicit mask is not provided,
then Nilearn's compute_epi_mask function will be used to derive a mask
from the first echo's data.
fittype : {'loglin', 'curvefit'}, optional
Monoexponential fitting method. 'loglin' uses the the default linear
fit to the log of the data. 'curvefit' uses a monoexponential fit to
the raw data, which is slightly slower but may be more accurate.
Default is 'loglin'.
combmode : {'t2s'}, optional
Combination scheme for TEs: 't2s' (Posse 1999, default).
tedpca : {'mdl', 'aic', 'kic', 'kundu', 'kundu-stabilize', float}, optional
Method with which to select components in TEDPCA.
If a float is provided, then it is assumed to represent percentage of variance
explained (0-1) to retain from PCA.
Default is 'mdl'.
tedort : :obj:`bool`, optional
Orthogonalize rejected components w.r.t. accepted ones prior to
denoising. Default is False.
gscontrol : {None, 'mir', 'gsr'} or :obj:`list`, optional
Perform additional denoising to remove spatially diffuse noise. Default
is None.
verbose : :obj:`bool`, optional
Generate intermediate and additional files. Default is False.
no_reports : obj:'bool', optional
Do not generate .html reports and .png plots. Default is false such
that reports are generated.
png_cmap : obj:'str', optional
Name of a matplotlib colormap to be used when generating figures.
Cannot be used with --no-png. Default is 'coolwarm'.
t2smap : :obj:`str`, optional
Precalculated T2* map in the same space as the input data. Values in
the map must be in seconds.
mixm : :obj:`str` or None, optional
File containing mixing matrix, to be used when re-running the workflow.
If not provided, ME-PCA and ME-ICA are done. Default is None.
ctab : :obj:`str` or None, optional
File containing component table from which to extract pre-computed
classifications, to be used with 'mixm' when re-running the workflow.
Default is None.
manacc : :obj:`list` of :obj:`int` or None, optional
List of manually accepted components. Can be a list of the components
numbers or None.
If provided, this parameter requires ``mixm`` and ``ctab`` to be provided as well.
Default is None.
Other Parameters
----------------
fixed_seed : :obj:`int`, optional
Value passed to ``mdp.numx_rand.seed()``.
Set to a positive integer value for reproducible ICA results;
otherwise, set to -1 for varying results across calls.
maxit : :obj:`int`, optional
Maximum number of iterations for ICA. Default is 500.
maxrestart : :obj:`int`, optional
Maximum number of attempts for ICA. If ICA fails to converge, the
fixed seed will be updated and ICA will be run again. If convergence
is achieved before maxrestart attempts, ICA will finish early.
Default is 10.
low_mem : :obj:`bool`, optional
Enables low-memory processing, including the use of IncrementalPCA.
May increase workflow duration. Default is False.
debug : :obj:`bool`, optional
Whether to run in debugging mode or not. Default is False.
quiet : :obj:`bool`, optional
If True, suppresses logging/printing of messages. Default is False.
Notes
-----
This workflow writes out several files. For a complete list of the files
generated by this workflow, please visit
https://tedana.readthedocs.io/en/latest/outputs.html
"""
out_dir = op.abspath(out_dir)
if not op.isdir(out_dir):
os.mkdir(out_dir)
# boilerplate
basename = 'report'
extension = 'txt'
repname = op.join(out_dir, (basename + '.' + extension))
repex = op.join(out_dir, (basename + '*'))
previousreps = glob(repex)
previousreps.sort(reverse=True)
for f in previousreps:
previousparts = op.splitext(f)
newname = previousparts[0] + '_old' + previousparts[1]
os.rename(f, newname)
refname = op.join(out_dir, '_references.txt')
# create logfile name
basename = 'tedana_'
extension = 'tsv'
start_time = datetime.datetime.now().strftime('%Y-%m-%dT%H%M%S')
logname = op.join(out_dir, (basename + start_time + '.' + extension))
utils.setup_loggers(logname, repname, refname, quiet=quiet, debug=debug)
LGR.info('Using output directory: {}'.format(out_dir))
# ensure tes are in appropriate format
tes = [float(te) for te in tes]
n_echos = len(tes)
# Coerce gscontrol to list
if not isinstance(gscontrol, list):
gscontrol = [gscontrol]
# Check value of tedpca *if* it is a float
tedpca = check_tedpca_value(tedpca, is_parser=False)
LGR.info('Loading input data: {}'.format([f for f in data]))
catd, ref_img = io.load_data(data, n_echos=n_echos)
io_generator = io.OutputGenerator(
ref_img,
convention=convention,
out_dir=out_dir,
prefix=prefix,
config="auto",
verbose=verbose,
)
n_samp, n_echos, n_vols = catd.shape
LGR.debug('Resulting data shape: {}'.format(catd.shape))
# check if TR is 0
img_t_r = io_generator.reference_img.header.get_zooms()[-1]
if img_t_r == 0:
raise IOError(
'Dataset has a TR of 0. This indicates incorrect'
' header information. To correct this, we recommend'
' using this snippet:'
'\n'
'https://gist.github.com/jbteves/032c87aeb080dd8de8861cb151bff5d6'
'\n'
'to correct your TR to the value it should be.')
if mixm is not None and op.isfile(mixm):
mixm = op.abspath(mixm)
# Allow users to re-run on same folder
mixing_name = io_generator.get_name("ICA mixing tsv")
if mixm != mixing_name:
shutil.copyfile(mixm, mixing_name)
shutil.copyfile(mixm,
op.join(io_generator.out_dir, op.basename(mixm)))
elif mixm is not None:
raise IOError('Argument "mixm" must be an existing file.')
if ctab is not None and op.isfile(ctab):
ctab = op.abspath(ctab)
# Allow users to re-run on same folder
metrics_name = io_generator.get_name("ICA metrics tsv")
if ctab != metrics_name:
shutil.copyfile(ctab, metrics_name)
shutil.copyfile(ctab,
op.join(io_generator.out_dir, op.basename(ctab)))
elif ctab is not None:
raise IOError('Argument "ctab" must be an existing file.')
if ctab and not mixm:
LGR.warning('Argument "ctab" requires argument "mixm".')
ctab = None
elif manacc is not None and (not mixm or not ctab):
LGR.warning('Argument "manacc" requires arguments "mixm" and "ctab".')
manacc = None
elif manacc is not None:
# coerce to list of integers
manacc = [int(m) for m in manacc]
if t2smap is not None and op.isfile(t2smap):
t2smap_file = io_generator.get_name('t2star img')
t2smap = op.abspath(t2smap)
# Allow users to re-run on same folder
if t2smap != t2smap_file:
shutil.copyfile(t2smap, t2smap_file)
elif t2smap is not None:
raise IOError('Argument "t2smap" must be an existing file.')
RepLGR.info("TE-dependence analysis was performed on input data.")
if mask and not t2smap:
# TODO: add affine check
LGR.info('Using user-defined mask')
RepLGR.info("A user-defined mask was applied to the data.")
elif t2smap and not mask:
LGR.info('Using user-defined T2* map to generate mask')
t2s_limited_sec = utils.load_image(t2smap)
t2s_limited = utils.sec2millisec(t2s_limited_sec)
t2s_full = t2s_limited.copy()
mask = (t2s_limited != 0).astype(int)
elif t2smap and mask:
LGR.info('Combining user-defined mask and T2* map to generate mask')
t2s_limited_sec = utils.load_image(t2smap)
t2s_limited = utils.sec2millisec(t2s_limited_sec)
t2s_full = t2s_limited.copy()
mask = utils.load_image(mask)
mask[t2s_limited == 0] = 0 # reduce mask based on T2* map
else:
LGR.info('Computing EPI mask from first echo')
first_echo_img = io.new_nii_like(io_generator.reference_img,
catd[:, 0, :])
mask = compute_epi_mask(first_echo_img)
RepLGR.info("An initial mask was generated from the first echo using "
"nilearn's compute_epi_mask function.")
# Create an adaptive mask with at least 1 good echo, for denoising
mask_denoise, masksum_denoise = utils.make_adaptive_mask(
catd,
mask=mask,
getsum=True,
threshold=1,
)
LGR.debug('Retaining {}/{} samples for denoising'.format(
mask_denoise.sum(), n_samp))
io_generator.save_file(masksum_denoise, "adaptive mask img")
# Create an adaptive mask with at least 3 good echoes, for classification
masksum_clf = masksum_denoise.copy()
masksum_clf[masksum_clf < 3] = 0
mask_clf = masksum_clf.astype(bool)
RepLGR.info(
"A two-stage masking procedure was applied, in which a liberal mask "
"(including voxels with good data in at least the first echo) was used for "
"optimal combination, T2*/S0 estimation, and denoising, while a more conservative mask "
"(restricted to voxels with good data in at least the first three echoes) was used for "
"the component classification procedure.")
LGR.debug('Retaining {}/{} samples for classification'.format(
mask_clf.sum(), n_samp))
if t2smap is None:
LGR.info('Computing T2* map')
t2s_limited, s0_limited, t2s_full, s0_full = decay.fit_decay(
catd, tes, mask_denoise, masksum_denoise, fittype)
# set a hard cap for the T2* map
# anything that is 10x higher than the 99.5 %ile will be reset to 99.5 %ile
cap_t2s = stats.scoreatpercentile(t2s_full.flatten(),
99.5,
interpolation_method='lower')
LGR.debug('Setting cap on T2* map at {:.5f}s'.format(
utils.millisec2sec(cap_t2s)))
t2s_full[t2s_full > cap_t2s * 10] = cap_t2s
io_generator.save_file(utils.millisec2sec(t2s_full), 't2star img')
io_generator.save_file(s0_full, 's0 img')
if verbose:
io_generator.save_file(utils.millisec2sec(t2s_limited),
'limited t2star img')
io_generator.save_file(s0_limited, 'limited s0 img')
# optimally combine data
data_oc = combine.make_optcom(catd,
tes,
masksum_denoise,
t2s=t2s_full,
combmode=combmode)
# regress out global signal unless explicitly not desired
if 'gsr' in gscontrol:
catd, data_oc = gsc.gscontrol_raw(catd, data_oc, n_echos, io_generator)
fout = io_generator.save_file(data_oc, 'combined img')
LGR.info('Writing optimally combined data set: {}'.format(fout))
if mixm is None:
# Identify and remove thermal noise from data
dd, n_components = decomposition.tedpca(catd,
data_oc,
combmode,
mask_clf,
masksum_clf,
t2s_full,
io_generator,
tes=tes,
algorithm=tedpca,
kdaw=10.,
rdaw=1.,
verbose=verbose,
low_mem=low_mem)
if verbose:
io_generator.save_file(utils.unmask(dd, mask_clf), 'whitened img')
# Perform ICA, calculate metrics, and apply decision tree
# Restart when ICA fails to converge or too few BOLD components found
keep_restarting = True
n_restarts = 0
seed = fixed_seed
while keep_restarting:
mmix, seed = decomposition.tedica(dd,
n_components,
seed,
maxit,
maxrestart=(maxrestart -
n_restarts))
seed += 1
n_restarts = seed - fixed_seed
# Estimate betas and compute selection metrics for mixing matrix
# generated from dimensionally reduced data using full data (i.e., data
# with thermal noise)
LGR.info(
'Making second component selection guess from ICA results')
required_metrics = [
'kappa', 'rho', 'countnoise', 'countsigFT2', 'countsigFS0',
'dice_FT2', 'dice_FS0', 'signal-noise_t', 'variance explained',
'normalized variance explained', 'd_table_score'
]
comptable = metrics.collect.generate_metrics(
catd,
data_oc,
mmix,
masksum_clf,
tes,
io_generator,
'ICA',
metrics=required_metrics,
)
comptable, metric_metadata = selection.kundu_selection_v2(
comptable, n_echos, n_vols)
n_bold_comps = comptable[comptable.classification ==
'accepted'].shape[0]
if (n_restarts < maxrestart) and (n_bold_comps == 0):
LGR.warning("No BOLD components found. Re-attempting ICA.")
elif (n_bold_comps == 0):
LGR.warning(
"No BOLD components found, but maximum number of restarts reached."
)
keep_restarting = False
else:
keep_restarting = False
RepLGR.disabled = True # Disable the report to avoid duplicate text
RepLGR.disabled = False # Re-enable the report after the while loop is escaped
else:
LGR.info('Using supplied mixing matrix from ICA')
mixing_file = io_generator.get_name("ICA mixing tsv")
mmix = pd.read_table(mixing_file).values
if ctab is None:
required_metrics = [
'kappa', 'rho', 'countnoise', 'countsigFT2', 'countsigFS0',
'dice_FT2', 'dice_FS0', 'signal-noise_t', 'variance explained',
'normalized variance explained', 'd_table_score'
]
comptable = metrics.collect.generate_metrics(
catd,
data_oc,
mmix,
masksum_clf,
tes,
io_generator,
'ICA',
metrics=required_metrics,
)
comptable, metric_metadata = selection.kundu_selection_v2(
comptable, n_echos, n_vols)
else:
comptable = pd.read_table(ctab)
if manacc is not None:
comptable, metric_metadata = selection.manual_selection(
comptable, acc=manacc)
# Write out ICA files.
comp_names = comptable["Component"].values
mixing_df = pd.DataFrame(data=mmix, columns=comp_names)
io_generator.save_file(mixing_df, "ICA mixing tsv")
betas_oc = utils.unmask(computefeats2(data_oc, mmix, mask_denoise),
mask_denoise)
io_generator.save_file(betas_oc, 'z-scored ICA components img')
# Save component table and associated json
io_generator.save_file(comptable, "ICA metrics tsv")
metric_metadata = metrics.collect.get_metadata(comptable)
io_generator.save_file(metric_metadata, "ICA metrics json")
decomp_metadata = {
"Method": ("Independent components analysis with FastICA "
"algorithm implemented by sklearn. "),
}
for comp_name in comp_names:
decomp_metadata[comp_name] = {
"Description":
"ICA fit to dimensionally-reduced optimally combined data.",
"Method": "tedana",
}
with open(io_generator.get_name("ICA decomposition json"), "w") as fo:
json.dump(decomp_metadata, fo, sort_keys=True, indent=4)
if comptable[comptable.classification == 'accepted'].shape[0] == 0:
LGR.warning('No BOLD components detected! Please check data and '
'results!')
mmix_orig = mmix.copy()
if tedort:
acc_idx = comptable.loc[~comptable.classification.str.
contains('rejected')].index.values
rej_idx = comptable.loc[comptable.classification.str.contains(
'rejected')].index.values
acc_ts = mmix[:, acc_idx]
rej_ts = mmix[:, rej_idx]
betas = np.linalg.lstsq(acc_ts, rej_ts, rcond=None)[0]
pred_rej_ts = np.dot(acc_ts, betas)
resid = rej_ts - pred_rej_ts
mmix[:, rej_idx] = resid
comp_names = [
io.add_decomp_prefix(comp,
prefix='ica',
max_value=comptable.index.max())
for comp in comptable.index.values
]
mixing_df = pd.DataFrame(data=mmix, columns=comp_names)
io_generator.save_file(mixing_df, "ICA orthogonalized mixing tsv")
RepLGR.info("Rejected components' time series were then "
"orthogonalized with respect to accepted components' time "
"series.")
io.writeresults(data_oc,
mask=mask_denoise,
comptable=comptable,
mmix=mmix,
n_vols=n_vols,
io_generator=io_generator)
if 'mir' in gscontrol:
gsc.minimum_image_regression(data_oc, mmix, mask_denoise, comptable,
io_generator)
if verbose:
io.writeresults_echoes(catd, mmix, mask_denoise, comptable,
io_generator)
# Write out BIDS-compatible description file
derivative_metadata = {
"Name":
"tedana Outputs",
"BIDSVersion":
"1.5.0",
"DatasetType":
"derivative",
"GeneratedBy": [{
"Name":
"tedana",
"Version":
__version__,
"Description":
("A denoising pipeline for the identification and removal "
"of non-BOLD noise from multi-echo fMRI data."),
"CodeURL":
"https://github.com/ME-ICA/tedana"
}]
}
with open(io_generator.get_name("data description json"), "w") as fo:
json.dump(derivative_metadata, fo, sort_keys=True, indent=4)
RepLGR.info("This workflow used numpy (Van Der Walt, Colbert, & "
"Varoquaux, 2011), scipy (Jones et al., 2001), pandas "
"(McKinney, 2010), scikit-learn (Pedregosa et al., 2011), "
"nilearn, and nibabel (Brett et al., 2019).")
RefLGR.info(
"Van Der Walt, S., Colbert, S. C., & Varoquaux, G. (2011). The "
"NumPy array: a structure for efficient numerical computation. "
"Computing in Science & Engineering, 13(2), 22.")
RefLGR.info("Jones E, Oliphant E, Peterson P, et al. SciPy: Open Source "
"Scientific Tools for Python, 2001-, http://www.scipy.org/")
RefLGR.info("McKinney, W. (2010, June). Data structures for statistical "
"computing in python. In Proceedings of the 9th Python in "
"Science Conference (Vol. 445, pp. 51-56).")
RefLGR.info("Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., "
"Thirion, B., Grisel, O., ... & Vanderplas, J. (2011). "
"Scikit-learn: Machine learning in Python. Journal of machine "
"learning research, 12(Oct), 2825-2830.")
RefLGR.info("Brett, M., Markiewicz, C. J., Hanke, M., Côté, M.-A., "
"Cipollini, B., McCarthy, P., … freec84. (2019, May 28). "
"nipy/nibabel. Zenodo. http://doi.org/10.5281/zenodo.3233118")
RepLGR.info("This workflow also used the Dice similarity index "
"(Dice, 1945; Sørensen, 1948).")
RefLGR.info("Dice, L. R. (1945). Measures of the amount of ecologic "
"association between species. Ecology, 26(3), 297-302.")
RefLGR.info(
"Sørensen, T. J. (1948). A method of establishing groups of "
"equal amplitude in plant sociology based on similarity of "
"species content and its application to analyses of the "
"vegetation on Danish commons. I kommission hos E. Munksgaard.")
with open(repname, 'r') as fo:
report = [line.rstrip() for line in fo.readlines()]
report = ' '.join(report)
with open(refname, 'r') as fo:
reference_list = sorted(list(set(fo.readlines())))
references = '\n'.join(reference_list)
report += '\n\nReferences:\n\n' + references
with open(repname, 'w') as fo:
fo.write(report)
if not no_reports:
LGR.info(
'Making figures folder with static component maps and timecourse plots.'
)
dn_ts, hikts, lowkts = io.denoise_ts(data_oc, mmix, mask_denoise,
comptable)
reporting.static_figures.carpet_plot(
optcom_ts=data_oc,
denoised_ts=dn_ts,
hikts=hikts,
lowkts=lowkts,
mask=mask_denoise,
io_generator=io_generator,
gscontrol=gscontrol,
)
reporting.static_figures.comp_figures(
data_oc,
mask=mask_denoise,
comptable=comptable,
mmix=mmix_orig,
io_generator=io_generator,
png_cmap=png_cmap,
)
if sys.version_info.major == 3 and sys.version_info.minor < 6:
warn_msg = ("Reports requested but Python version is less than "
"3.6.0. Dynamic reports will not be generated.")
LGR.warn(warn_msg)
else:
LGR.info('Generating dynamic report')
reporting.generate_report(io_generator, tr=img_t_r)
LGR.info('Workflow completed')
utils.teardown_loggers()
os.remove(refname)
def tedpca(data_cat, data_oc, combmode, mask, t2s, t2sG,
ref_img, tes, algorithm='mdl', source_tes=-1, kdaw=10., rdaw=1.,
out_dir='.', verbose=False, low_mem=False):
"""
Use principal components analysis (PCA) to identify and remove thermal
noise from multi-echo data.
Parameters
----------
data_cat : (S x E x T) array_like
Input functional data
data_oc : (S x T) array_like
Optimally combined time series data
combmode : {'t2s', 'paid'} str
How optimal combination of echos should be made, where 't2s' indicates
using the method of Posse 1999 and 'paid' indicates using the method of
Poser 2006
mask : (S,) array_like
Boolean mask array
t2s : (S,) array_like
Map of voxel-wise T2* estimates.
t2sG : (S,) array_like
Map of voxel-wise T2* estimates.
ref_img : :obj:`str` or img_like
Reference image to dictate how outputs are saved to disk
tes : :obj:`list`
List of echo times associated with `data_cat`, in milliseconds
algorithm : {'mle', 'kundu', 'kundu-stabilize', 'mdl', 'aic', 'kic'}, optional
Method with which to select components in TEDPCA. Default is 'mdl'. PCA
decomposition with the mdl, kic and aic options are based on a Moving Average
(stationary Gaussian) process and are ordered from most to least aggresive.
See (Li et al., 2007).
source_tes : :obj:`int` or :obj:`list` of :obj:`int`, optional
Which echos to use in PCA. Values -1 and 0 are special, where a value
of -1 will indicate using the optimal combination of the echos
and 0 will indicate using all the echos. A list can be provided
to indicate a subset of echos.
Default: -1
kdaw : :obj:`float`, optional
Dimensionality augmentation weight for Kappa calculations. Must be a
non-negative float, or -1 (a special value). Default is 10.
rdaw : :obj:`float`, optional
Dimensionality augmentation weight for Rho calculations. Must be a
non-negative float, or -1 (a special value). Default is 1.
out_dir : :obj:`str`, optional
Output directory.
verbose : :obj:`bool`, optional
Whether to output files from fitmodels_direct or not. Default: False
low_mem : :obj:`bool`, optional
Whether to use incremental PCA (for low-memory systems) or not.
Default: False
Returns
-------
kept_data : (S x T) :obj:`numpy.ndarray`
Dimensionally reduced optimally combined functional data
n_components : :obj:`int`
Number of components retained from PCA decomposition
Notes
-----
====================== =================================================
Notation Meaning
====================== =================================================
:math:`\\kappa` Component pseudo-F statistic for TE-dependent
(BOLD) model.
:math:`\\rho` Component pseudo-F statistic for TE-independent
(artifact) model.
:math:`v` Voxel
:math:`V` Total number of voxels in mask
:math:`\\zeta` Something
:math:`c` Component
:math:`p` Something else
====================== =================================================
Steps:
1. Variance normalize either multi-echo or optimally combined data,
depending on settings.
2. Decompose normalized data using PCA or SVD.
3. Compute :math:`{\\kappa}` and :math:`{\\rho}`:
.. math::
{\\kappa}_c = \\frac{\\sum_{v}^V {\\zeta}_{c,v}^p * \
F_{c,v,R_2^*}}{\\sum {\\zeta}_{c,v}^p}
{\\rho}_c = \\frac{\\sum_{v}^V {\\zeta}_{c,v}^p * \
F_{c,v,S_0}}{\\sum {\\zeta}_{c,v}^p}
4. Some other stuff. Something about elbows.
5. Classify components as thermal noise if they meet both of the
following criteria:
- Nonsignificant :math:`{\\kappa}` and :math:`{\\rho}`.
- Nonsignificant variance explained.
Outputs:
This function writes out several files:
====================== =================================================
Filename Content
====================== =================================================
pca_decomposition.json PCA component table.
pca_mixing.tsv PCA mixing matrix.
pca_components.nii.gz Component weight maps.
====================== =================================================
"""
if low_mem and algorithm == 'mle':
LGR.warning('Low memory option is not compatible with MLE '
'dimensionality estimation. Switching to Kundu decision '
'tree.')
algorithm = 'kundu'
if algorithm == 'mle':
alg_str = "using MLE dimensionality estimation (Minka, 2001)"
RefLGR.info("Minka, T. P. (2001). Automatic choice of dimensionality "
"for PCA. In Advances in neural information processing "
"systems (pp. 598-604).")
elif algorithm == 'kundu':
alg_str = ("followed by the Kundu component selection decision "
"tree (Kundu et al., 2013)")
RefLGR.info("Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192.")
elif algorithm == 'kundu-stabilize':
alg_str = ("followed by the 'stabilized' Kundu component "
"selection decision tree (Kundu et al., 2013)")
RefLGR.info("Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192.")
else:
alg_str = ("based on the PCA component estimation with a Moving Average"
"(stationary Gaussian) process (Li et al., 2007)")
RefLGR.info("Li, Y.O., Adalı, T. and Calhoun, V.D., (2007). "
"Estimating the number of independent components for "
"functional magnetic resonance imaging data. "
"Human brain mapping, 28(11), pp.1251-1266.")
if source_tes == -1:
dat_str = "the optimally combined data"
elif source_tes == 0:
dat_str = "the z-concatenated multi-echo data"
else:
dat_str = "a z-concatenated subset of echoes from the input data"
RepLGR.info("Principal component analysis {0} was applied to "
"{1} for dimensionality reduction.".format(alg_str, dat_str))
n_samp, n_echos, n_vols = data_cat.shape
source_tes = np.array([int(ee) for ee in str(source_tes).split(',')])
if len(source_tes) == 1 and source_tes[0] == -1:
LGR.info('Computing PCA of optimally combined multi-echo data')
data = data_oc[mask, :][:, np.newaxis, :]
elif len(source_tes) == 1 and source_tes[0] == 0:
LGR.info('Computing PCA of spatially concatenated multi-echo data')
data = data_cat[mask, ...]
else:
LGR.info('Computing PCA of echo #{0}'.format(','.join([str(ee) for ee in source_tes])))
data = np.stack([data_cat[mask, ee, :] for ee in source_tes - 1], axis=1)
eim = np.squeeze(_utils.eimask(data))
data = np.squeeze(data[eim])
data_z = ((data.T - data.T.mean(axis=0)) / data.T.std(axis=0)).T # var normalize ts
data_z = (data_z - data_z.mean()) / data_z.std() # var normalize everything
if algorithm in ['mdl', 'aic', 'kic']:
data_img = io.new_nii_like(
ref_img, utils.unmask(utils.unmask(data, eim), mask))
mask_img = io.new_nii_like(ref_img,
utils.unmask(eim, mask).astype(int))
voxel_comp_weights, varex, varex_norm, comp_ts = ma_pca.ma_pca(
data_img, mask_img, algorithm)
elif algorithm == 'mle':
voxel_comp_weights, varex, varex_norm, comp_ts = run_mlepca(data_z)
elif low_mem:
voxel_comp_weights, varex, comp_ts = low_mem_pca(data_z)
varex_norm = varex / varex.sum()
else:
ppca = PCA(copy=False, n_components=(n_vols - 1))
ppca.fit(data_z)
comp_ts = ppca.components_.T
varex = ppca.explained_variance_
voxel_comp_weights = np.dot(np.dot(data_z, comp_ts),
np.diag(1. / varex))
varex_norm = varex / varex.sum()
# Compute Kappa and Rho for PCA comps
eimum = np.atleast_2d(eim)
eimum = np.transpose(eimum, np.argsort(eimum.shape)[::-1])
eimum = eimum.prod(axis=1)
o = np.zeros((mask.shape[0], *eimum.shape[1:]))
o[mask, ...] = eimum
eimum = np.squeeze(o).astype(bool)
# Normalize each component's time series
vTmixN = stats.zscore(comp_ts, axis=0)
comptable, _, _, _ = metrics.dependence_metrics(data_cat,
data_oc,
comp_ts,
t2s,
tes,
ref_img,
reindex=False,
mmixN=vTmixN,
algorithm=None,
label='mepca_',
out_dir=out_dir,
verbose=verbose)
# varex_norm from PCA overrides varex_norm from dependence_metrics,
# but we retain the original
comptable['estimated normalized variance explained'] = \
comptable['normalized variance explained']
comptable['normalized variance explained'] = varex_norm
# write component maps to 4D image
comp_ts_z = stats.zscore(comp_ts, axis=0)
comp_maps = utils.unmask(computefeats2(data_oc, comp_ts_z, mask), mask)
io.filewrite(comp_maps, op.join(out_dir, 'pca_components.nii.gz'), ref_img)
# Select components using decision tree
if algorithm == 'kundu':
comptable = kundu_tedpca(comptable, n_echos, kdaw, rdaw, stabilize=False)
elif algorithm == 'kundu-stabilize':
comptable = kundu_tedpca(comptable, n_echos, kdaw, rdaw, stabilize=True)
elif algorithm == 'mle':
LGR.info('Selected {0} components with MLE dimensionality '
'detection'.format(comptable.shape[0]))
comptable['classification'] = 'accepted'
comptable['rationale'] = ''
elif algorithm in ['mdl', 'aic', 'kic']:
LGR.info('Selected {0} components with {1} dimensionality '
'detection'.format(comptable.shape[0], algorithm))
comptable['classification'] = 'accepted'
comptable['rationale'] = ''
# Save decomposition
comp_names = [io.add_decomp_prefix(comp, prefix='pca', max_value=comptable.index.max())
for comp in comptable.index.values]
mixing_df = pd.DataFrame(data=comp_ts, columns=comp_names)
mixing_df.to_csv(op.join(out_dir, 'pca_mixing.tsv'), sep='\t', index=False)
data_type = 'optimally combined data' if source_tes == -1 else 'z-concatenated data'
comptable['Description'] = 'PCA fit to {0}.'.format(data_type)
mmix_dict = {}
mmix_dict['Method'] = ('Principal components analysis implemented by '
'sklearn. Components are sorted by variance '
'explained in descending order. '
'Component signs are flipped to best match the '
'data.')
io.save_comptable(comptable, op.join(out_dir, 'pca_decomposition.json'),
label='pca', metadata=mmix_dict)
acc = comptable[comptable.classification == 'accepted'].index.values
n_components = acc.size
voxel_kept_comp_weighted = (voxel_comp_weights[:, acc] * varex[None, acc])
kept_data = np.dot(voxel_kept_comp_weighted, comp_ts[:, acc].T)
kept_data = stats.zscore(kept_data, axis=1) # variance normalize time series
kept_data = stats.zscore(kept_data, axis=None) # variance normalize everything
return kept_data, n_components
def tedpca(
data_cat,
data_oc,
combmode,
mask,
adaptive_mask,
t2sG,
io_generator,
tes,
algorithm="aic",
kdaw=10.0,
rdaw=1.0,
verbose=False,
low_mem=False,
):
"""
Use principal components analysis (PCA) to identify and remove thermal
noise from multi-echo data.
Parameters
----------
data_cat : (S x E x T) array_like
Input functional data
data_oc : (S x T) array_like
Optimally combined time series data
combmode : {'t2s', 'paid'} str
How optimal combination of echos should be made, where 't2s' indicates
using the method of Posse 1999 and 'paid' indicates using the method of
Poser 2006
mask : (S,) array_like
Boolean mask array
adaptive_mask : (S,) array_like
Array where each value indicates the number of echoes with good signal
for that voxel. This mask may be thresholded; for example, with values
less than 3 set to 0.
For more information on thresholding, see `make_adaptive_mask`.
t2sG : (S,) array_like
Map of voxel-wise T2* estimates.
io_generator : :obj:`tedana.io.OutputGenerator`
The output generation object for this workflow
tes : :obj:`list`
List of echo times associated with `data_cat`, in milliseconds
algorithm : {'kundu', 'kundu-stabilize', 'mdl', 'aic', 'kic', float}, optional
Method with which to select components in TEDPCA. PCA
decomposition with the mdl, kic and aic options are based on a Moving Average
(stationary Gaussian) process and are ordered from most to least aggressive
(see Li et al., 2007).
If a float is provided, then it is assumed to represent percentage of variance
explained (0-1) to retain from PCA.
If an int is provided, then it is assumed to be the number of components
to select
Default is 'aic'.
kdaw : :obj:`float`, optional
Dimensionality augmentation weight for Kappa calculations. Must be a
non-negative float, or -1 (a special value). Default is 10.
rdaw : :obj:`float`, optional
Dimensionality augmentation weight for Rho calculations. Must be a
non-negative float, or -1 (a special value). Default is 1.
verbose : :obj:`bool`, optional
Whether to output files from fitmodels_direct or not. Default: False
low_mem : :obj:`bool`, optional
Whether to use incremental PCA (for low-memory systems) or not.
This is only compatible with the "kundu" or "kundu-stabilize" algorithms.
Default: False
Returns
-------
kept_data : (S x T) :obj:`numpy.ndarray`
Dimensionally reduced optimally combined functional data
n_components : :obj:`int`
Number of components retained from PCA decomposition
Notes
-----
====================== =================================================
Notation Meaning
====================== =================================================
:math:`\\kappa` Component pseudo-F statistic for TE-dependent
(BOLD) model.
:math:`\\rho` Component pseudo-F statistic for TE-independent
(artifact) model.
:math:`v` Voxel
:math:`V` Total number of voxels in mask
:math:`\\zeta` Something
:math:`c` Component
:math:`p` Something else
====================== =================================================
Steps:
1. Variance normalize either multi-echo or optimally combined data,
depending on settings.
2. Decompose normalized data using PCA or SVD.
3. Compute :math:`{\\kappa}` and :math:`{\\rho}`:
.. math::
{\\kappa}_c = \\frac{\\sum_{v}^V {\\zeta}_{c,v}^p * \
F_{c,v,R_2^*}}{\\sum {\\zeta}_{c,v}^p}
{\\rho}_c = \\frac{\\sum_{v}^V {\\zeta}_{c,v}^p * \
F_{c,v,S_0}}{\\sum {\\zeta}_{c,v}^p}
4. Some other stuff. Something about elbows.
5. Classify components as thermal noise if they meet both of the
following criteria:
- Nonsignificant :math:`{\\kappa}` and :math:`{\\rho}`.
- Nonsignificant variance explained.
Outputs:
This function writes out several files:
=========================== =============================================
Default Filename Content
=========================== =============================================
desc-PCA_metrics.tsv PCA component table
desc-PCA_metrics.json Metadata sidecar file describing the
component table
desc-PCA_mixing.tsv PCA mixing matrix
desc-PCA_components.nii.gz Component weight maps
desc-PCA_decomposition.json Metadata sidecar file describing the PCA
decomposition
=========================== =============================================
See Also
--------
:func:`tedana.utils.make_adaptive_mask` : The function used to create
the ``adaptive_mask`` parameter.
:py:mod:`tedana.constants` : The module describing the filenames for
various naming conventions
"""
if algorithm == "kundu":
alg_str = "followed by the Kundu component selection decision tree (Kundu et al., 2013)"
RefLGR.info(
"Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192."
)
elif algorithm == "kundu-stabilize":
alg_str = (
"followed by the 'stabilized' Kundu component "
"selection decision tree (Kundu et al., 2013)"
)
RefLGR.info(
"Kundu, P., Brenowitz, N. D., Voon, V., Worbe, Y., "
"Vértes, P. E., Inati, S. J., ... & Bullmore, E. T. "
"(2013). Integrated strategy for improving functional "
"connectivity mapping using multiecho fMRI. Proceedings "
"of the National Academy of Sciences, 110(40), "
"16187-16192."
)
elif isinstance(algorithm, Number):
if isinstance(algorithm, float):
alg_str = (
"in which the number of components was determined based on a "
"variance explained threshold"
)
else:
alg_str = "in which the number of components is pre-defined"
else:
alg_str = (
"based on the PCA component estimation with a Moving Average"
"(stationary Gaussian) process (Li et al., 2007)"
)
RefLGR.info(
"Li, Y.O., Adalı, T. and Calhoun, V.D., (2007). "
"Estimating the number of independent components for "
"functional magnetic resonance imaging data. "
"Human brain mapping, 28(11), pp.1251-1266."
)
RepLGR.info(
"Principal component analysis {0} was applied to "
"the optimally combined data for dimensionality "
"reduction.".format(alg_str)
)
n_samp, n_echos, n_vols = data_cat.shape
LGR.info(
f"Computing PCA of optimally combined multi-echo data with selection criteria: {algorithm}"
)
data = data_oc[mask, :]
data_z = ((data.T - data.T.mean(axis=0)) / data.T.std(axis=0)).T # var normalize ts
data_z = (data_z - data_z.mean()) / data_z.std() # var normalize everything
if algorithm in ["mdl", "aic", "kic"]:
data_img = io.new_nii_like(io_generator.reference_img, utils.unmask(data, mask))
mask_img = io.new_nii_like(io_generator.reference_img, mask.astype(int))
ma_pca = MovingAveragePCA(criterion=algorithm, normalize=True)
_ = ma_pca.fit_transform(data_img, mask_img)
# Extract results from maPCA
voxel_comp_weights = ma_pca.u_
varex = ma_pca.explained_variance_
varex_norm = ma_pca.explained_variance_ratio_
comp_ts = ma_pca.components_.T
aic = ma_pca.aic_
kic = ma_pca.kic_
mdl = ma_pca.mdl_
varex_90 = ma_pca.varexp_90_
varex_95 = ma_pca.varexp_95_
all_comps = ma_pca.all_
# Extract number of components and variance explained for logging and plotting
n_aic = aic["n_components"]
aic_varexp = np.round(aic["explained_variance_total"], 3)
n_kic = kic["n_components"]
kic_varexp = np.round(kic["explained_variance_total"], 3)
n_mdl = mdl["n_components"]
mdl_varexp = np.round(mdl["explained_variance_total"], 3)
n_varex_90 = varex_90["n_components"]
varex_90_varexp = np.round(varex_90["explained_variance_total"], 3)
n_varex_95 = varex_95["n_components"]
varex_95_varexp = np.round(varex_95["explained_variance_total"], 3)
all_varex = np.round(all_comps["explained_variance_total"], 3)
# Print out the results
LGR.info("Optimal number of components based on different criteria:")
LGR.info(
f"AIC: {n_aic} | KIC: {n_kic} | MDL: {n_mdl} | 90% varexp: {n_varex_90} "
f"| 95% varexp: {n_varex_95}"
)
LGR.info("Explained variance based on different criteria:")
LGR.info(
f"AIC: {aic_varexp}% | KIC: {kic_varexp}% | MDL: {mdl_varexp}% | "
f"90% varexp: {varex_90_varexp}% | 95% varexp: {varex_95_varexp}%"
)
pca_optimization_curves = np.array([aic["value"], kic["value"], mdl["value"]])
pca_criteria_components = np.array(
[
n_aic,
n_kic,
n_mdl,
n_varex_90,
n_varex_95,
]
)
# Plot maPCA optimization curves
LGR.info("Plotting maPCA optimization curves")
plot_pca_results(pca_optimization_curves, pca_criteria_components, all_varex, io_generator)
# Save maPCA results into a dictionary
mapca_results = {
"aic": {
"n_components": n_aic,
"explained_variance_total": aic_varexp,
"curve": aic["value"],
},
"kic": {
"n_components": n_kic,
"explained_variance_total": kic_varexp,
"curve": kic["value"],
},
"mdl": {
"n_components": n_mdl,
"explained_variance_total": mdl_varexp,
"curve": mdl["value"],
},
"varex_90": {
"n_components": n_varex_90,
"explained_variance_total": varex_90_varexp,
},
"varex_95": {
"n_components": n_varex_95,
"explained_variance_total": varex_95_varexp,
},
}
# Save dictionary
io_generator.save_file(mapca_results, "PCA cross component metrics json")
elif isinstance(algorithm, Number):
ppca = PCA(copy=False, n_components=algorithm, svd_solver="full")
ppca.fit(data_z)
comp_ts = ppca.components_.T
varex = ppca.explained_variance_
voxel_comp_weights = np.dot(np.dot(data_z, comp_ts), np.diag(1.0 / varex))
varex_norm = ppca.explained_variance_ratio_
elif low_mem:
voxel_comp_weights, varex, varex_norm, comp_ts = low_mem_pca(data_z)
else:
ppca = PCA(copy=False, n_components=(n_vols - 1))
ppca.fit(data_z)
comp_ts = ppca.components_.T
varex = ppca.explained_variance_
voxel_comp_weights = np.dot(np.dot(data_z, comp_ts), np.diag(1.0 / varex))
varex_norm = ppca.explained_variance_ratio_
# Compute Kappa and Rho for PCA comps
required_metrics = [
"kappa",
"rho",
"countnoise",
"countsigFT2",
"countsigFS0",
"dice_FT2",
"dice_FS0",
"signal-noise_t",
"variance explained",
"normalized variance explained",
"d_table_score",
]
comptable = metrics.collect.generate_metrics(
data_cat,
data_oc,
comp_ts,
adaptive_mask,
tes,
io_generator,
"PCA",
metrics=required_metrics,
)
# varex_norm from PCA overrides varex_norm from dependence_metrics,
# but we retain the original
comptable["estimated normalized variance explained"] = comptable[
"normalized variance explained"
]
comptable["normalized variance explained"] = varex_norm
# write component maps to 4D image
comp_maps = utils.unmask(computefeats2(data_oc, comp_ts, mask), mask)
io_generator.save_file(comp_maps, "z-scored PCA components img")
# Select components using decision tree
if algorithm == "kundu":
comptable, metric_metadata = kundu_tedpca(
comptable,
n_echos,
kdaw,
rdaw,
stabilize=False,
)
elif algorithm == "kundu-stabilize":
comptable, metric_metadata = kundu_tedpca(
comptable,
n_echos,
kdaw,
rdaw,
stabilize=True,
)
else:
if isinstance(algorithm, float):
alg_str = "variance explained-based"
elif isinstance(algorithm, int):
alg_str = "a fixed number of components and no"
else:
alg_str = algorithm
LGR.info(
f"Selected {comptable.shape[0]} components with {round(100*varex_norm.sum(),2)}% "
f"normalized variance explained using {alg_str} dimensionality detection"
)
comptable["classification"] = "accepted"
comptable["rationale"] = ""
# Save decomposition files
comp_names = [
io.add_decomp_prefix(comp, prefix="pca", max_value=comptable.index.max())
for comp in comptable.index.values
]
mixing_df = pd.DataFrame(data=comp_ts, columns=comp_names)
io_generator.save_file(mixing_df, "PCA mixing tsv")
# Save component table and associated json
io_generator.save_file(comptable, "PCA metrics tsv")
metric_metadata = metrics.collect.get_metadata(comptable)
io_generator.save_file(metric_metadata, "PCA metrics json")
decomp_metadata = {
"Method": (
"Principal components analysis implemented by sklearn. "
"Components are sorted by variance explained in descending order. "
),
}
for comp_name in comp_names:
decomp_metadata[comp_name] = {
"Description": "PCA fit to optimally combined data.",
"Method": "tedana",
}
io_generator.save_file(decomp_metadata, "PCA decomposition json")
acc = comptable[comptable.classification == "accepted"].index.values
n_components = acc.size
voxel_kept_comp_weighted = voxel_comp_weights[:, acc] * varex[None, acc]
kept_data = np.dot(voxel_kept_comp_weighted, comp_ts[:, acc].T)
kept_data = stats.zscore(kept_data, axis=1) # variance normalize time series
kept_data = stats.zscore(kept_data, axis=None) # variance normalize everything
return kept_data, n_components
def fitmodels_direct(catd,
mmix,
mask,
t2s,
t2s_full,
tes,
combmode,
ref_img,
reindex=False,
mmixN=None,
full_sel=True,
label=None,
out_dir='.',
verbose=False):
"""
Fit TE-dependence and -independence models to components.
Parameters
----------
catd : (S x E x T) array_like
Input data, where `S` is samples, `E` is echos, and `T` is time
mmix : (T x C) array_like
Mixing matrix for converting input data to component space, where `C`
is components and `T` is the same as in `catd`
mask : (S [x E]) array_like
Boolean mask array
t2s : (S [x T]) array_like
Limited T2* map or timeseries.
t2s_full : (S [x T]) array_like
Full T2* map or timeseries. For voxels with good signal in only one
echo, which are zeros in the limited T2* map, this map uses the T2*
estimate using the first two echoes.
tes : list
List of echo times associated with `catd`, in milliseconds
combmode : {'t2s', 'paid'} str
How optimal combination of echos should be made, where 't2s' indicates
using the method of Posse 1999 and 'paid' indicates using the method of
Poser 2006
ref_img : str or img_like
Reference image to dictate how outputs are saved to disk
reindex : bool, optional
Default: False
mmixN : array_like, optional
Default: None
full_sel : bool, optional
Whether to perform selection of components based on Rho/Kappa scores.
Default: True
Returns
-------
seldict : dict
comptable : (C x X) :obj:`pandas.DataFrame`
Component metric table. One row for each component, with a column for
each metric. The index is the component number.
betas : :obj:`numpy.ndarray`
mmix_new : :obj:`numpy.ndarray`
"""
if not (catd.shape[0] == t2s.shape[0] == t2s_full.shape[0] ==
mask.shape[0]):
raise ValueError('First dimensions (number of samples) of catd ({0}), '
't2s ({1}), and mask ({2}) do not '
'match'.format(catd.shape[0], t2s.shape[0],
mask.shape[0]))
elif catd.shape[1] != len(tes):
raise ValueError('Second dimension of catd ({0}) does not match '
'number of echoes provided (tes; '
'{1})'.format(catd.shape[1], len(tes)))
elif catd.shape[2] != mmix.shape[0]:
raise ValueError('Third dimension (number of volumes) of catd ({0}) '
'does not match first dimension of '
'mmix ({1})'.format(catd.shape[2], mmix.shape[0]))
elif t2s.shape != t2s_full.shape:
raise ValueError('Shape of t2s array {0} does not match shape of '
't2s_full array {1}'.format(t2s.shape,
t2s_full.shape))
elif t2s.ndim == 2:
if catd.shape[2] != t2s.shape[1]:
raise ValueError('Third dimension (number of volumes) of catd '
'({0}) does not match second dimension of '
't2s ({1})'.format(catd.shape[2], t2s.shape[1]))
mask = t2s != 0 # Override mask because problems
# compute optimal combination of raw data
tsoc = combine.make_optcom(catd,
tes,
mask,
t2s=t2s_full,
combmode=combmode,
verbose=False).astype(float)[mask]
# demean optimal combination
tsoc_dm = tsoc - tsoc.mean(axis=-1, keepdims=True)
# compute un-normalized weight dataset (features)
if mmixN is None:
mmixN = mmix
WTS = computefeats2(utils.unmask(tsoc, mask), mmixN, mask, normalize=False)
# compute PSC dataset - shouldn't have to refit data
tsoc_B = get_coeffs(tsoc_dm, mmix, mask=None)
tsoc_Babs = np.abs(tsoc_B)
PSC = tsoc_B / tsoc.mean(axis=-1, keepdims=True) * 100
# compute skews to determine signs based on unnormalized weights,
# correct mmix & WTS signs based on spatial distribution tails
signs = stats.skew(WTS, axis=0)
signs /= np.abs(signs)
mmix = mmix.copy()
mmix *= signs
WTS *= signs
PSC *= signs
totvar = (tsoc_B**2).sum()
totvar_norm = (WTS**2).sum()
# compute Betas and means over TEs for TE-dependence analysis
betas = get_coeffs(catd, mmix,
np.repeat(mask[:, np.newaxis], len(tes), axis=1))
n_samp, n_echos, n_components = betas.shape
n_voxels = mask.sum()
n_data_voxels = (t2s != 0).sum()
mu = catd.mean(axis=-1, dtype=float)
tes = np.reshape(tes, (n_echos, 1))
fmin, _, _ = utils.getfbounds(n_echos)
# mask arrays
mumask = mu[t2s != 0]
t2smask = t2s[t2s != 0]
betamask = betas[t2s != 0]
# set up Xmats
X1 = mumask.T # Model 1
X2 = np.tile(tes, (1, n_data_voxels)) * mumask.T / t2smask.T # Model 2
# tables for component selection
kappas = np.zeros([n_components])
rhos = np.zeros([n_components])
varex = np.zeros([n_components])
varex_norm = np.zeros([n_components])
Z_maps = np.zeros([n_voxels, n_components])
F_R2_maps = np.zeros([n_data_voxels, n_components])
F_S0_maps = np.zeros([n_data_voxels, n_components])
Z_clmaps = np.zeros([n_voxels, n_components])
F_R2_clmaps = np.zeros([n_data_voxels, n_components])
F_S0_clmaps = np.zeros([n_data_voxels, n_components])
Br_R2_clmaps = np.zeros([n_voxels, n_components])
Br_S0_clmaps = np.zeros([n_voxels, n_components])
pred_R2_maps = np.zeros([n_data_voxels, n_echos, n_components])
pred_S0_maps = np.zeros([n_data_voxels, n_echos, n_components])
LGR.info('Fitting TE- and S0-dependent models to components')
for i_comp in range(n_components):
# size of B is (n_echoes, n_samples)
B = np.atleast_3d(betamask)[:, :, i_comp].T
alpha = (np.abs(B)**2).sum(axis=0)
varex[i_comp] = (tsoc_B[:, i_comp]**2).sum() / totvar * 100.
varex_norm[i_comp] = (utils.unmask(WTS, mask)[t2s != 0][:, i_comp]**2).sum() /\
totvar_norm * 100.
# S0 Model
# (S,) model coefficient map
coeffs_S0 = (B * X1).sum(axis=0) / (X1**2).sum(axis=0)
pred_S0 = X1 * np.tile(coeffs_S0, (n_echos, 1))
pred_S0_maps[:, :, i_comp] = pred_S0.T
SSE_S0 = (B - pred_S0)**2
SSE_S0 = SSE_S0.sum(axis=0) # (S,) prediction error map
F_S0 = (alpha - SSE_S0) * (n_echos - 1) / (SSE_S0)
F_S0_maps[:, i_comp] = F_S0
# R2 Model
coeffs_R2 = (B * X2).sum(axis=0) / (X2**2).sum(axis=0)
pred_R2 = X2 * np.tile(coeffs_R2, (n_echos, 1))
pred_R2_maps[:, :, i_comp] = pred_R2.T
SSE_R2 = (B - pred_R2)**2
SSE_R2 = SSE_R2.sum(axis=0)
F_R2 = (alpha - SSE_R2) * (n_echos - 1) / (SSE_R2)
F_R2_maps[:, i_comp] = F_R2
# compute weights as Z-values
wtsZ = (WTS[:, i_comp] - WTS[:, i_comp].mean()) / WTS[:, i_comp].std()
wtsZ[np.abs(wtsZ) > Z_MAX] = (
Z_MAX * (np.abs(wtsZ) / wtsZ))[np.abs(wtsZ) > Z_MAX]
Z_maps[:, i_comp] = wtsZ
# compute Kappa and Rho
F_S0[F_S0 > F_MAX] = F_MAX
F_R2[F_R2 > F_MAX] = F_MAX
norm_weights = np.abs(
np.squeeze(utils.unmask(wtsZ, mask)[t2s != 0]**2.))
kappas[i_comp] = np.average(F_R2, weights=norm_weights)
rhos[i_comp] = np.average(F_S0, weights=norm_weights)
# tabulate component values
comptable = np.vstack([kappas, rhos, varex, varex_norm]).T
if reindex:
# re-index all components in Kappa order
sort_idx = comptable[:, 0].argsort()[::-1]
comptable = comptable[sort_idx, :]
mmix_new = mmix[:, sort_idx]
betas = betas[..., sort_idx]
pred_R2_maps = pred_R2_maps[:, :, sort_idx]
pred_S0_maps = pred_S0_maps[:, :, sort_idx]
F_S0_maps = F_S0_maps[:, sort_idx]
F_R2_maps = F_R2_maps[:, sort_idx]
Z_maps = Z_maps[:, sort_idx]
WTS = WTS[:, sort_idx]
PSC = PSC[:, sort_idx]
tsoc_B = tsoc_B[:, sort_idx]
tsoc_Babs = tsoc_Babs[:, sort_idx]
else:
mmix_new = mmix
if verbose:
# Echo-specific weight maps for each of the ICA components.
io.filewrite(betas, op.join(out_dir,
'{0}betas_catd.nii'.format(label)),
ref_img)
# Echo-specific maps of predicted values for R2 and S0 models for each
# component.
io.filewrite(utils.unmask(pred_R2_maps, mask),
op.join(out_dir, '{0}R2_pred.nii'.format(label)), ref_img)
io.filewrite(utils.unmask(pred_S0_maps, mask),
op.join(out_dir, '{0}S0_pred.nii'.format(label)), ref_img)
# Weight maps used to average metrics across voxels
io.filewrite(utils.unmask(Z_maps**2., mask),
op.join(out_dir, '{0}metric_weights.nii'.format(label)),
ref_img)
comptable = pd.DataFrame(comptable,
columns=[
'kappa', 'rho', 'variance explained',
'normalized variance explained'
])
comptable.index.name = 'component'
# full selection including clustering criteria
seldict = None
if full_sel:
LGR.info('Performing spatial clustering of components')
csize = np.max([int(n_voxels * 0.0005) + 5, 20])
LGR.debug('Using minimum cluster size: {}'.format(csize))
for i_comp in range(n_components):
# Cluster-extent threshold and binarize F-maps
ccimg = io.new_nii_like(
ref_img,
np.squeeze(utils.unmask(F_R2_maps[:, i_comp], t2s != 0)))
F_R2_clmaps[:,
i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=fmin,
mask=mask,
binarize=True)
countsigFR2 = F_R2_clmaps[:, i_comp].sum()
ccimg = io.new_nii_like(
ref_img,
np.squeeze(utils.unmask(F_S0_maps[:, i_comp], t2s != 0)))
F_S0_clmaps[:,
i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=fmin,
mask=mask,
binarize=True)
countsigFS0 = F_S0_clmaps[:, i_comp].sum()
# Cluster-extent threshold and binarize Z-maps with CDT of p < 0.05
ccimg = io.new_nii_like(
ref_img, np.squeeze(utils.unmask(Z_maps[:, i_comp], t2s != 0)))
Z_clmaps[:, i_comp] = utils.threshold_map(ccimg,
min_cluster_size=csize,
threshold=1.95,
mask=mask,
binarize=True)
# Cluster-extent threshold and binarize ranked signal-change map
ccimg = io.new_nii_like(
ref_img,
utils.unmask(stats.rankdata(tsoc_Babs[:, i_comp]), t2s != 0))
Br_R2_clmaps[:, i_comp] = utils.threshold_map(
ccimg,
min_cluster_size=csize,
threshold=(max(tsoc_Babs.shape) - countsigFR2),
mask=mask,
binarize=True)
Br_S0_clmaps[:, i_comp] = utils.threshold_map(
ccimg,
min_cluster_size=csize,
threshold=(max(tsoc_Babs.shape) - countsigFS0),
mask=mask,
binarize=True)
seldict = {}
selvars = [
'WTS', 'tsoc_B', 'PSC', 'Z_maps', 'F_R2_maps', 'F_S0_maps',
'Z_clmaps', 'F_R2_clmaps', 'F_S0_clmaps', 'Br_R2_clmaps',
'Br_S0_clmaps'
]
for vv in selvars:
seldict[vv] = eval(vv)
return seldict, comptable, betas, mmix_new
def threshold_to_match(maps, n_sig_voxels, mask, ref_img, csize=None):
"""Cluster-extent threshold a map to target number of significant voxels.
Resulting maps have roughly the requested number of significant voxels, after cluster-extent
thresholding.
Parameters
----------
maps : (M x C) array_like
Statistical maps to be thresholded.
n_sig_voxels : (C) array_like
Number of significant voxels to threshold to, for each map in maps.
mask : (S) array_like
Binary mask.
ref_img : img_like
Reference image to convert to niimgs with.
csize : :obj:`int` or :obj:`None`, optional
Minimum cluster size. If None, standard thresholding (non-cluster-extent) will be done.
Default is None.
Returns
-------
clmaps : (S x C) array_like
Cluster-extent thresholded and binarized maps.
"""
assert maps.shape[1] == n_sig_voxels.shape[0]
n_voxels, n_components = maps.shape
abs_maps = np.abs(maps)
if csize is None:
csize = np.max([int(n_voxels * 0.0005) + 5, 20])
else:
csize = int(csize)
clmaps = np.zeros([n_voxels, n_components], bool)
for i_comp in range(n_components):
# Initial cluster-defining threshold is defined based on the number
# of significant voxels from the F-statistic maps. This threshold
# will be relaxed until the number of significant voxels from both
# maps is roughly equal.
ccimg = io.new_nii_like(ref_img, utils.unmask(stats.rankdata(abs_maps[:, i_comp]), mask))
step = int(n_sig_voxels[i_comp] / 10)
rank_thresh = n_voxels - n_sig_voxels[i_comp]
while True:
clmap = utils.threshold_map(
ccimg,
min_cluster_size=csize,
threshold=rank_thresh,
mask=mask,
binarize=True,
)
if rank_thresh <= 0: # all voxels significant
break
diff = n_sig_voxels[i_comp] - clmap.sum()
if diff < 0 or clmap.sum() == 0:
rank_thresh += step
clmap = utils.threshold_map(
ccimg,
min_cluster_size=csize,
threshold=rank_thresh,
mask=mask,
binarize=True,
)
break
else:
rank_thresh -= step
clmaps[:, i_comp] = clmap
return clmaps
