def _write_image(dat,
fname,
bids=False,
mat=torch.eye(4),
file=None,
dtype='float32'):
""" Write data to nifti.
"""
if bids:
p, n = os.path.split(fname)
s = n.split('_')
fname = os.path.join(p, '_'.join(s[:-1] + ['space-unires'] + [s[-1]]))
savef(dat, fname, like=file, affine=mat)
def _write_output(dat, mat, file=None, prefix='', odir=None, nam=None):
"""Write preprocessed output to disk.
"""
if odir is not None:
os.makedirs(odir, exist_ok=True)
pth = []
for n in range(len(dat)):
if file[n] is not None:
filename = file[n].filename()
else:
filename = 'nitorch_file'
pth.append(file_mod(filename, odir=odir, prefix=prefix, nam=nam))
savef(dat[n], pth[n], like=file[n], affine=mat[n])
if len(dat) == 1:
pth = pth[0]
return pth
def do_apply(fnames, phi, jac):
"""Correct files with a given polarity"""
for fname in fnames:
dir, base, ext = py.fileparts(fname)
ofname = options.output
ofname = ofname.format(dir=dir or '.', sep=os.sep, base=base,
ext=ext)
if options.verbose:
print(f'unwarp {fname} \n'
f' -> {ofname}')
f = io.map(fname)
d = f.fdata(device=device)
d = utils.movedim(d, readout, -1)
d = _deform1d(d, phi)
if jac is not None:
d *= jac
d = utils.movedim(d, -1, readout)
io.savef(d, ofname, like=fname)
def _cli(args):
"""Command-line interface for `smooth` without exception handling"""
args = args or sys.argv[1:]
options = parser(args)
if options.help:
print(help)
return
fwhm = options.fwhm
unit = 'mm'
if isinstance(fwhm[-1], str):
*fwhm, unit = fwhm
fwhm = make_list(fwhm, 3)
options.output = make_list(options.output, len(options.files))
for fname, ofname in zip(options.files, options.output):
f = io.map(fname)
vx = voxel_size(f.affine).tolist()
dim = len(vx)
if unit == 'mm':
fwhm1 = [f / v for f, v in zip(fwhm, vx)]
else:
fwhm1 = fwhm[:len(vx)]
dat = f.fdata()
dat = movedim_front2back(dat, dim)
dat = smooth(dat,
type=options.method,
fwhm=fwhm1,
basis=options.basis,
bound=options.padding,
dim=dim)
dat = movedim_back2front(dat, dim)
folder, base, ext = fileparts(fname)
ofname = ofname.format(dir=folder or '.',
base=base,
ext=ext,
sep=os.path.sep)
io.savef(dat, ofname, like=f)
def _main(options):
if isinstance(options.gpu, str):
device = torch.device(options.gpu)
else:
assert isinstance(options.gpu, int)
device = torch.device(f'cuda:{options.gpu}')
if not torch.cuda.is_available():
device = 'cpu'
# prepare options
estatics_opt = ESTATICSOptions()
estatics_opt.likelihood = options.likelihood
estatics_opt.verbose = options.verbose >= 1
estatics_opt.plot = options.verbose >= 2
estatics_opt.recon.space = options.space
if isinstance(options.space, str) and options.space != 'mean':
for c, contrast in enumerate(options.contrast):
if contrast.name == options.space:
estatics_opt.recon.space = c
break
estatics_opt.backend.device = device
estatics_opt.optim.nb_levels = options.levels
estatics_opt.optim.max_iter_rls = options.iter
estatics_opt.optim.tolerance = options.tol
estatics_opt.regularization.norm = options.regularization
estatics_opt.regularization.factor = [*options.lam_intercept, options.lam_decay]
estatics_opt.distortion.enable = options.meetup
estatics_opt.distortion.bending = options.lam_meetup
estatics_opt.preproc.register = options.register
# prepare files
contrasts = []
distortion = []
for i, c in enumerate(options.contrast):
# read meta-parameters
meta = {}
if c.te:
te, unit = c.te, ''
if isinstance(te[-1], str):
*te, unit = te
if unit:
if unit == 'ms':
te = [t * 1e-3 for t in te]
elif unit not in ('s', 'sec'):
raise ValueError(f'TE unit: {unit}')
if c.echo_spacing:
delta, *unit = c.echo_spacing
unit = unit[0] if unit else ''
if unit == 'ms':
delta = delta * 1e-3
elif unit not in ('s', 'sec'):
raise ValueError(f'echo spacing unit: {unit}')
ne = sum(io.map(f).unsqueeze(-1).shape[3] for f in c.echoes)
te = [te[0] + e*delta for e in range(ne)]
meta['te'] = te
# map volumes
contrasts.append(qio.GradientEchoMulti.from_fname(c.echoes, **meta))
if c.readout:
layout = spatial.affine_to_layout(contrasts[-1].affine)
layout = spatial.volume_layout_to_name(layout)
readout = None
for j, l in enumerate(layout):
if l.lower() in c.readout.lower():
readout = j - 3
contrasts[-1].readout = readout
if c.b0:
bw = c.bandwidth
b0, *unit = c.b0
unit = unit[-1] if unit else 'vx'
fb0 = b0.map(b0)
b0 = fb0.fdata(device=device)
b0 = spatial.reslice(b0, fb0.affine, contrasts[-1][0].affine,
contrasts[-1][0].shape)
if unit.lower() == 'hz':
if not bw:
raise ValueError('Bandwidth required to convert fieldmap'
'from Hz to voxel')
b0 /= bw
b0 = DenseDistortion(b0)
distortion.append(b0)
else:
distortion.append(None)
# run algorithm
[te0, r2s, *b0] = estatics(contrasts, distortion, opt=estatics_opt)
# write results
# --- intercepts ---
odir0 = options.odir
for i, te1 in enumerate(te0):
ifname = contrasts[i].echo(0).volume.fname
odir, obase, oext = py.fileparts(ifname)
odir = odir0 or odir
obase = obase + '_TE0'
ofname = os.path.join(odir, obase + oext)
io.savef(te1.volume, ofname, affine=te1.affine, like=ifname, te=0, dtype='float32')
# --- decay ---
ifname = contrasts[0].echo(0).volume.fname
odir, obase, oext = py.fileparts(ifname)
odir = odir0 or odir
io.savef(r2s.volume, os.path.join(odir, 'R2star' + oext), affine=r2s.affine, dtype='float32')
# --- fieldmap + undistorted ---
if b0:
b0 = b0[0]
for i, b01 in enumerate(b0):
ifname = contrasts[i].echo(0).volume.fname
odir, obase, oext = py.fileparts(ifname)
odir = odir0 or odir
obase = obase + '_B0'
ofname = os.path.join(odir, obase + oext)
io.savef(b01.volume, ofname, affine=b01.affine, like=ifname, te=0, dtype='float32')
for i, (c, b) in enumerate(zip(contrasts, b0)):
readout = c.readout
grid_up, grid_down, jac_up, jac_down = b.exp2(
add_identity=True, jacobian=True)
for j, e in enumerate(c):
blip = e.blip or (2*(j % 2) - 1)
grid_blip = grid_down if blip > 0 else grid_up # inverse of
jac_blip = jac_down if blip > 0 else jac_up # forward model
ifname = e.volume.fname
odir, obase, oext = py.fileparts(ifname)
odir = odir0 or odir
obase = obase + '_unwrapped'
ofname = os.path.join(odir, obase + oext)
d = e.fdata(device=device)
d, _ = pull1d(d, grid_blip, readout)
d *= jac_blip
io.savef(d, ofname, affine=e.affine, like=ifname)
del d
del grid_up, grid_down, jac_up, jac_down
if options.register:
for i, c in enumerate(contrasts):
for j, e in enumerate(c):
ifname = e.volume.fname
odir, obase, oext = py.fileparts(ifname)
odir = odir0 or odir
obase = obase + '_registered'
ofname = os.path.join(odir, obase + oext)
io.save(e.volume, ofname, affine=e.affine)
def _warp_image(option,
affine=None,
nonlin=None,
dim=None,
device=None,
odir=None):
"""Warp and save the moving and fixed images from a loss object"""
if not (option.mov.output or option.mov.resliced or option.fix.output
or option.fix.resliced):
return
fix, fix_affine = _map_image(option.fix.files, dim=dim)
mov, mov_affine = _map_image(option.mov.files, dim=dim)
fix_affine = fix_affine.float()
mov_affine = mov_affine.float()
dim = dim or (fix.dim - 1)
if option.fix.world: # overwrite orientation matrix
fix_affine = io.transforms.map(option.fix.world).fdata().squeeze()
for transform in (option.fix.affine or []):
transform = io.transforms.map(transform).fdata().squeeze()
fix_affine = spatial.affine_lmdiv(transform, fix_affine)
if option.mov.world: # overwrite orientation matrix
mov_affine = io.transforms.map(option.mov.world).fdata().squeeze()
for transform in (option.mov.affine or []):
transform = io.transforms.map(transform).fdata().squeeze()
mov_affine = spatial.affine_lmdiv(transform, mov_affine)
# moving
if option.mov.output or option.mov.resliced:
ifname = option.mov.files[0]
idir, base, ext = py.fileparts(ifname)
odir_mov = odir or idir or '.'
image = objects.Image(mov.fdata(rand=True, device=device),
dim=dim,
affine=mov_affine,
bound=option.mov.bound,
extrapolate=option.mov.extrapolate)
if option.mov.output:
target_affine = mov_affine
target_shape = image.shape
if affine and affine.position[0].lower() in 'ms':
aff = affine.exp(recompute=False, cache_result=True)
target_affine = spatial.affine_lmdiv(aff, target_affine)
fname = option.mov.output.format(dir=odir_mov,
base=base,
sep=os.path.sep,
ext=ext)
print(f'Minimal reslice: {ifname} -> {fname} ...', end=' ')
warped = _warp_image1(image,
target_affine,
target_shape,
affine=affine,
nonlin=nonlin)
io.savef(warped, fname, like=ifname, affine=target_affine)
print('done.')
del warped
if option.mov.resliced:
target_affine = fix_affine
target_shape = fix.shape[1:]
fname = option.mov.resliced.format(dir=odir_mov,
base=base,
sep=os.path.sep,
ext=ext)
print(f'Full reslice: {ifname} -> {fname} ...', end=' ')
warped = _warp_image1(image,
target_affine,
target_shape,
affine=affine,
nonlin=nonlin,
reslice=True)
io.savef(warped, fname, like=ifname, affine=target_affine)
print('done.')
del warped
# fixed
if option.fix.output or option.fix.resliced:
ifname = option.fix.files[0]
idir, base, ext = py.fileparts(ifname)
odir_fix = odir or idir or '.'
image = objects.Image(fix.fdata(rand=True, device=device),
dim=dim,
affine=fix_affine,
bound=option.fix.bound,
extrapolate=option.fix.extrapolate)
if option.fix.output:
target_affine = fix_affine
target_shape = image.shape
if affine and affine.position[0].lower() in 'fs':
aff = affine.exp(recompute=False, cache_result=True)
target_affine = spatial.affine_matmul(aff, target_affine)
fname = option.fix.output.format(dir=odir_fix,
base=base,
sep=os.path.sep,
ext=ext)
print(f'Minimal reslice: {ifname} -> {fname} ...', end=' ')
warped = _warp_image1(image,
target_affine,
target_shape,
affine=affine,
nonlin=nonlin,
backward=True)
io.savef(warped, fname, like=ifname, affine=target_affine)
print('done.')
del warped
if option.fix.resliced:
target_affine = mov_affine
target_shape = mov.shape[1:]
fname = option.fix.resliced.format(dir=odir_fix,
base=base,
sep=os.path.sep,
ext=ext)
print(f'Full reslice: {ifname} -> {fname} ...', end=' ')
warped = _warp_image1(image,
target_affine,
target_shape,
affine=affine,
nonlin=nonlin,
backward=True,
reslice=True)
io.savef(warped, fname, like=ifname, affine=target_affine)
print('done.')
del warped
def _main(options):
device = setup_device(*options.device)
dim = 3
# ------------------------------------------------------------------
# COMPUTE PYRAMID
# ------------------------------------------------------------------
pyramids = _prepare_pyramid_levels(options.loss, options.pyramid, dim)
# ------------------------------------------------------------------
# BUILD LOSSES
# ------------------------------------------------------------------
loss_list, image_dict = _build_losses(options, pyramids, device)
can_use_2nd_order = all(loss.loss.order >= 2 for loss in loss_list)
# ------------------------------------------------------------------
# BUILD AFFINE
# ------------------------------------------------------------------
affine, affine_optim = _build_affine(options, can_use_2nd_order)
# ------------------------------------------------------------------
# BUILD DENSE
# ------------------------------------------------------------------
nonlin, nonlin_optim = _build_nonlin(options, can_use_2nd_order, affine,
image_dict)
if not affine and not nonlin:
raise ValueError('At least one of @affine or @nonlin must be used.')
# ------------------------------------------------------------------
# BACKEND STUFF
# ------------------------------------------------------------------
if options.verbose > 1:
import matplotlib
matplotlib.use('TkAgg')
# local losses may benefit from selecting the best conv
torch.backends.cudnn.enabled = True
torch.backends.cudnn.benchmark = True
# ------------------------------------------------------------------
# PERFORM REGISTRATION
# ------------------------------------------------------------------
_do_register(loss_list, affine, nonlin, affine_optim, nonlin_optim,
options)
# ------------------------------------------------------------------
# WRITE RESULTS
# ------------------------------------------------------------------
if affine:
affine = affine[-1]
if affine and options.affine.output:
odir = options.odir or py.fileparts(
options.loss[0].fix.files[0])[0] or '.'
fname = options.affine.output.format(dir=odir,
sep=os.path.sep,
name=options.affine.name)
print('Affine ->', fname)
aff = affine.exp(cache_result=True, recompute=False)
io.transforms.savef(aff.cpu(), fname, type=1) # 1 = RAS_TO_RAS
if nonlin and options.nonlin.output:
odir = options.odir or py.fileparts(
options.loss[0].fix.files[0])[0] or '.'
fname = options.nonlin.output.format(dir=odir,
sep=os.path.sep,
name=options.nonlin.name)
io.savef(nonlin.dat.dat, fname, affine=nonlin.affine)
print('Nonlin ->', fname)
for loss in options.loss:
_warp_image(loss,
affine=affine,
nonlin=nonlin,
dim=dim,
device=device,
odir=options.odir)
def denoise_mri(*dat_x,
affine_x=None,
lam_scl=opt.lam_scl,
lr=opt.learning_rate,
max_iter=opt.max_iter,
tolerance=opt.tolerance,
verbose=opt.verbose,
device=opt.device,
do_write=opt.do_write,
dir_out=opt.dir_out):
"""Denoises a multi-channel MR image by solving:
dat_y_hat = 0.5*sum_c(tau_c*sum_i((dat_x_ci - dat_y_ci)^2)) + jtv(dat_y_1, ..., dat_y_C; lam)
using PyTorch's auto-diff.
If input is given as paths to files, outputs prefixed 'den_' is written based on options
'dir_out' and 'do_write'.
Reference:
Brudfors, Mikael, et al. "MRI super-resolution using multi-channel total variation."
Annual Conference on Medical Image Understanding and Analysis. Springer, Cham, 2018.
Parameters
----------
dat_x : (nchannels, dmx, dmy, dmz) tensor or sequence[str]
Input noisy image data
affine_x : (4, 4) tensor, optional
Input images' affine matrix. If not given, assumes identity.
lam_scl : float, default=10.0
Scaling of regularisation values
lr : float, default=1e1
Optimiser learning rate
max_iter : int, default=10000
Maximum number of fitting iterations
tolerance : float, default=1e-8
Convergence threshold (when to stop iterating)
verbose : bool, default=True
Print to terminal?
device : torch.device, default='cuda'
Torch device
do_write : bool, default=True
If input is given as paths to files, output is written to disk,
prefixed 'den_' to 'dir_out'
dir_out : str, optional
Directory where to write output, default is same as input.
Returns
----------
dat_y_hat : (nchannels, dmx, dmy, dmz) tensor
Denoised image data
"""
# read data from disk
if isinstance(dat_x,(list, tuple)) and \
sum(isinstance(dat_x[i],str) for i in range(len(dat_x))) == len(dat_x):
dat_x, affine_x, nii = _get_image_data(dat_x, device=device)
else:
do_write = False # input is tensor, do not write to disk
# backend
device = dat_x.device
dtype = dat_x.dtype
# estimate hyper-parameters
tau = torch.zeros(dat_x.shape[0], device=device, dtype=dtype)
lam = torch.zeros(dat_x.shape[0], device=device, dtype=dtype)
for i in range(dat_x.shape[0]):
prm0, prm1 = estimate_noise(dat_x[i, ...], show_fit=False)
sd_bg = prm0['sd']
mean_fg = prm1['mean']
tau[i] = 1 / sd_bg.float()**2
lam[i] = math.sqrt(1 / dat_x.shape[0]) / mean_fg.float(
) # modulates with number of channels (as in JTV reg)
# print("tau={:}".format(tau))
# print("lam={:}".format(lam))
# affine matrices
if affine_x is None:
affine_x = torch.eye(4, device=device, dtype=dtype)
# voxel size
vx = voxel_size(affine_x)
# scale regularisation
lam = lam_scl * lam
# initial estimate of reconstruction
dat_y_hat = torch.zeros_like(dat_x)
dat_y_hat = torch.nn.Parameter(dat_y_hat, requires_grad=True)
# prepare optimiser and scheduler
optim = torch.optim.Adam([dat_y_hat], lr=lr) # Adam
# optim = torch.optim.SGD([dat_y_hat], lr=lr, momentum=0.9) # SGD
scheduler = ReduceLROnPlateau(optim)
# optimisation loop
loss_vals = torch.zeros(max_iter + 1, dtype=torch.float64)
cnt_conv = 0
for n_iter in range(1, max_iter + 1):
# set gradients to zero (PyTorch accumulates the gradients on subsequent backward passes)
optim.zero_grad()
# compute reconstruction loss
loss_val = _loss_ssqd_jtv(dat_x, dat_y_hat, tau, lam, vx=vx)
# differentiate reconstruction loss w.r.t. dat_y_hat
loss_val.backward()
# store loss
loss_vals[n_iter] = loss_val.item()
# update reconstruction
optim.step()
# compute gain
gain = get_gain(loss_vals[:n_iter + 1], monotonicity='decreasing')
if verbose:
# print to screen
with torch.no_grad():
if n_iter % 25 == 0:
print('n_iter={:4d}, loss={:12.6f}, gain={:0.10}, lr={:g}'. \
format(n_iter, loss_val.item(), gain, optim.param_groups[0]['lr']), end='\n') # end='\r'
if n_iter > 10 and gain.abs() < tolerance:
cnt_conv += 1
if cnt_conv == 5:
# finished
break
else:
cnt_conv = 0
# incorporate scheduler
if scheduler is not None and n_iter % 10 == 0:
if isinstance(scheduler, ReduceLROnPlateau):
scheduler.step(loss_val)
else:
scheduler.step()
if do_write:
# write output to disk
if dir_out is not None:
os.makedirs(dir_out, exist_ok=True)
for i in range(dat_y_hat.shape[0]):
fname = file_replace(nii.fname,
prefix='den_',
dir=dir_out,
suffix='_' + str(i))
savef(dat_y_hat[i, ...], fname, like=nii)
return dat_y_hat
def reslice(moving, fname, like, inv=False, lin=None, nonlin=None,
interpolation=1, bound='dct2', extrapolate=False, device=None,
verbose=True):
"""Apply the linear and non-linear components of the transform and
reslice the image to the target space.
Notes
-----
.. The shape and general orientation of the moving image is kept untouched.
.. The linear transform is composed with the original orientation matrix.
.. The non-linear component is "wrapped" in the input space, where it
is applied.
.. This function writes a new file (it does not modify the input files
in place).
Parameters
----------
moving : ImageFile
An object describing a moving image.
fname : list of str
Output filename for each input file of the moving image
(since images can be encoded over multiple volumes)
like : ImageFile
An object describing the target space
inv : bool, default=False
True if we are warping the fixed image to the moving space.
In the case, `moving` should be a `FixedImageFile` and `like` a
`MovingImageFile`. Else it should be a `MovingImageFile` and `'like`
a `FixedImageFile`.
lin : (4, 4) tensor, optional
Linear (or rather affine) transformation
nonlin : dict, optional
Non-linear displacement field, with keys:
disp : (..., 3) tensor
Displacement field (in voxels)
affine : (4, 4) tensor
Orientation matrix of the displacement field
interpolation : int, default=1
bound : str, default='dct2'
extrapolate : bool, default = False
device : default='cpu'
"""
nonlin = nonlin or dict(disp=None, affine=None)
prm = dict(interpolation=interpolation, bound=bound, extrapolate=extrapolate)
moving_affine = moving.affine.to(device)
fixed_affine = like.affine.to(device)
if inv:
# affine-corrected fixed space
if lin is not None:
fix2lin = affine_matmul(lin, fixed_affine)
else:
fix2lin = fixed_affine
if nonlin['disp'] is not None:
# fixed voxels to param voxels (warps param to fixed)
fix2nlin = affine_lmdiv(nonlin['affine'].to(device), fix2lin)
if samespace(fix2nlin, nonlin['disp'].shape[:-1], like.shape):
g = smalldef(nonlin['disp'].to(device))
else:
g = affine_grid(fix2nlin, like.shape)
g += pull_grid(nonlin['disp'].to(device), g)
# param to moving
nlin2mov = affine_lmdiv(moving_affine, nonlin['affine'].to(device))
g = affine_matvec(nlin2mov, g)
else:
g = affine_lmdiv(moving_affine, fix2lin)
g = affine_grid(g, like.shape)
else:
# affine-corrected moving space
if lin is not None:
mov2nlin = affine_matmul(lin, moving_affine)
else:
mov2nlin = moving_affine
if nonlin['disp'] is not None:
# fixed voxels to param voxels (warps param to fixed)
fix2nlin = affine_lmdiv(nonlin['affine'].to(device), fixed_affine)
if samespace(fix2nlin, nonlin['disp'].shape[:-1], like.shape):
g = smalldef(nonlin['disp'].to(device))
else:
g = affine_grid(fix2nlin, like.shape)
g += pull_grid(nonlin['disp'].to(device), g)
# param voxels to moving voxels (warps moving to fixed)
nlin2mov = affine_lmdiv(mov2nlin, nonlin['affine'].to(device))
g = affine_matvec(nlin2mov, g)
else:
g = affine_lmdiv(mov2nlin, fixed_affine)
g = affine_grid(g, like.shape)
if moving.type == 'labels':
prm['interpolation'] = 0
for file, ofname in zip(moving.files, fname):
if verbose:
print(f'Resliced: {file.fname}\n'
f' -> {ofname}')
dat = io.volumes.loadf(file.fname, rand=True, device=device)
dat = dat.reshape([*file.shape, file.channels])
if g is not None:
dat = utils.movedim(dat, -1, 0)
dat = pull(dat, g, **prm)
dat = utils.movedim(dat, 0, -1)
io.savef(dat.cpu(), ofname, like=file.fname, affine=like.affine.cpu())
def main(options):
# find readout direction
f = io.map(options.echoes[0])
affine, shape = f.affine, f.shape
readout = get_readout(options.direction, affine, shape, options.verbose)
if not options.reversed:
reversed_echoes = options.synth
else:
reversed_echoes = options.reversed
# do EPIC
fit = epic(options.echoes,
reverse_echoes=reversed_echoes,
fieldmap=options.fieldmap,
extrapolate=options.extrapolate,
bandwidth=options.bandwidth,
polarity=options.polarity,
readout=readout,
slicewise=options.slicewise,
lam=options.penalty,
max_iter=options.maxiter,
tol=options.tolerance,
verbose=options.verbose,
device=get_device(options.gpu))
# save volumes
input, output = options.echoes, options.output
if len(output) != len(input):
if len(output) == 1:
if '{base}' in output[0]:
output = [output[0]] * len(input)
elif len(output) != len(fit):
raise ValueError(f'There should be either one output file, '
f'or as many output files as input files, '
f'or as many output files as echoes. Got '
f'{len(output)} output files, {len(input)} '
f'input files, and {len(fit)} echoes.')
if len(output) == 1:
dir, base, ext = py.fileparts(input[0])
output = output[0]
if '{n}' in output:
for n, echo in enumerate(fit):
out = output.format(dir=dir,
sep=os.sep,
base=base,
ext=ext,
n=n)
io.savef(echo, out, like=input[0])
else:
output = output.format(dir=dir, sep=os.sep, base=base, ext=ext)
io.savef(torch.movedim(fit, 0, -1), output, like=input[0])
elif len(output) == len(input):
for i, (inp, out) in enumerate(zip(input, output)):
dir, base, ext = py.fileparts(inp)
out = out.format(dir=dir, sep=os.sep, base=base, ext=ext, n=i)
ne = [*io.map(inp).shape, 1][3]
io.savef(fit[:ne].movedim(0, -1), out, like=inp)
fit = fit[ne:]
else:
assert len(output) == len(fit)
dir, base, ext = py.fileparts(input[0])
for n, (echo, out) in enumerate(zip(fit, output)):
out = out.format(dir=dir, sep=os.sep, base=base, ext=ext, n=n)
io.savef(echo, out, like=input[0])
def main_fit(options):
"""
Estimate a displacement field from opposite polarity images
"""
device = get_device(options.gpu)
# map input files
f0 = io.map(options.pos_file)
f1 = io.map(options.neg_file)
dim = f0.affine.shape[-1] - 1
# map mask
fm = None
if options.mask:
fm = io.map(options.mask)
# detect readout direction
readout = get_readout(options.readout, f0.affine, f0.shape[-dim:])
# detect penalty
penalty_type = 'bending'
penalties = options.penalty
if penalties and isinstance(penalties[-1], str):
*penalties, penalty_type = penalties
if not penalties:
penalties = [1]
if penalty_type[0] == 'b':
penalty_type = 'bending'
elif penalty_type[0] == 'm':
penalty_type = 'membrane'
else:
raise ValueError('Unknown penalty type', penalty_type)
downs = options.downsample
max_iter = options.max_iter
tolerance = options.tolerance
nb_levels = max(len(penalties), len(max_iter), len(tolerance), len(downs))
penalties = py.make_list(penalties, nb_levels)
tolerance = py.make_list(tolerance, nb_levels)
max_iter = py.make_list(max_iter, nb_levels)
downs = py.make_list(downs, nb_levels)
# load
d00 = f0.fdata(device='cpu')
d11 = f1.fdata(device='cpu')
dmask = fm.fdata(device='cpu') if fm else None
# fit
vel = mask = None
aff = last_aff = f0.affine
last_dwn = None
for penalty, n, tol, dwn in zip(penalties, max_iter, tolerance, downs):
if dwn != last_dwn:
d0, aff = downsample(d00.to(device), f0.affine, dwn)
d1, _ = downsample(d11.to(device), f1.affine, dwn)
vx = spatial.voxel_size(aff)
if vel is not None:
vel = upsample_vel(vel, last_aff, aff, d0.shape[-dim:], readout)
last_aff = aff
if fm:
mask, _ = downsample(dmask.to(device), f1.affine, dwn)
last_dwn = dwn
scl = py.prod(d00.shape) / py.prod(d0.shape)
penalty = penalty * scl
kernel = get_kernel(options.kernel, aff, d0.shape[-dim:], dwn)
# prepare loss
if options.loss == 'mse':
prm0, _ = estimate_noise(d0)
prm1, _ = estimate_noise(d1)
sd = ((prm0['sd'].log() + prm1['sd'].log())/2).exp()
print(sd.item())
loss = MSE(lam=1/(sd*sd), dim=dim)
elif options.loss == 'lncc':
loss = LNCC(dim=dim, patch=kernel)
elif options.loss == 'lgmm':
if options.bins == 1:
loss = LNCC(dim=dim, patch=kernel)
else:
loss = LGMMH(dim=dim, patch=kernel, bins=options.bins)
elif options.loss == 'gmm':
if options.bins == 1:
loss = NCC(dim=dim)
else:
loss = GMMH(dim=dim, bins=options.bins)
else:
loss = NCC(dim=dim)
# fit
vel = topup_fit(d0, d1, loss=loss, dim=readout, vx=vx, ndim=dim,
model=('svf' if options.diffeo else 'smalldef'),
lam=penalty, penalty=penalty_type, vel=vel,
modulation=options.modulation, max_iter=n,
tolerance=tol, verbose=options.verbose, mask=mask)
del d0, d1, d00, d11
# upsample
vel = upsample_vel(vel, aff, f0.affine, f0.shape[-dim:], readout)
# save
dir, base, ext = py.fileparts(options.pos_file)
fname = options.output
fname = fname.format(dir=dir or '.', sep=os.sep, base=base, ext=ext)
io.savef(vel, fname, like=options.pos_file, dtype='float32')
def savef(self, fname, *args, **kwargs):
"""Save to disk"""
io.savef(self.volume, fname, *args, **kwargs)
def write_outputs(z, prm, options):
# prepare filenames
ref_native = options.input[0]
ref_mni = options.tpm[0] if options.tpm else path_spm_prior()
format_dict = get_format_dict(ref_native, options.output)
# move channels to back
backend = utils.backend(z)
if (options.nobias_nat or options.nobias_mni or options.nobias_wrp
or options.all_nat or options.all_mni or options.all_wrp):
dat, _, affine = get_data(options.input, options.mask, None, 3,
**backend)
# --- native space -------------------------------------------------
if options.prob_nat or options.all_nat:
fname = options.prob_nat or '{dir}{sep}{base}.prob.nat{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('prob.nat ->', fname)
io.savef(torch.movedim(z, 0, -1),
fname,
like=ref_native,
dtype='float32')
if options.labels_nat or options.all_nat:
fname = options.labels_nat or '{dir}{sep}{base}.labels.nat{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('labels.nat ->', fname)
io.save(z.argmax(0), fname, like=ref_native, dtype='int16')
if (options.bias_nat or options.all_nat) and options.bias:
bias = prm['bias']
fname = options.bias_nat or '{dir}{sep}{base}.bias.nat{ext}'
if len(options.input) == 1:
fname = fname.format(**format_dict)
if options.verbose > 0:
print('bias.nat ->', fname)
io.savef(torch.movedim(bias, 0, -1),
fname,
like=ref_native,
dtype='float32')
else:
for c, (bias1, ref1) in enumerate(zip(bias, options.input)):
format_dict1 = get_format_dict(ref1, options.output)
fname = fname.format(**format_dict1)
if options.verbose > 0:
print(f'bias.nat.{c+1} ->', fname)
io.savef(bias1, fname, like=ref1, dtype='float32')
del bias
if (options.nobias_nat or options.all_nat) and options.bias:
nobias = dat * prm['bias']
fname = options.nobias_nat or '{dir}{sep}{base}.nobias.nat{ext}'
if len(options.input) == 1:
fname = fname.format(**format_dict)
if options.verbose > 0:
print('nobias.nat ->', fname)
io.savef(torch.movedim(nobias, 0, -1), fname, like=ref_native)
else:
for c, (nobias1, ref1) in enumerate(zip(bias, options.input)):
format_dict1 = get_format_dict(ref1, options.output)
fname = fname.format(**format_dict1)
if options.verbose > 0:
print(f'nobias.nat.{c+1} ->', fname)
io.savef(nobias1, fname, like=ref1)
del nobias
if (options.warp_nat or options.all_nat) and options.warp:
warp = prm['warp']
fname = options.warp_nat or '{dir}{sep}{base}.warp.nat{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('warp.nat ->', fname)
io.savef(warp, fname, like=ref_native, dtype='float32')
# --- MNI space ----------------------------------------------------
if options.tpm is False:
# No template -> no MNI space
return
fref = io.map(ref_mni)
mni_affine, mni_shape = fref.affine, fref.shape[:3]
dat_affine = io.map(ref_native).affine
mni_affine = mni_affine.to(**backend)
dat_affine = dat_affine.to(**backend)
prm_affine = prm['affine'].to(**backend)
dat_affine = prm_affine @ dat_affine
if options.mni_vx:
vx = spatial.voxel_size(mni_affine)
scl = vx / options.mni_vx
mni_affine, mni_shape = spatial.affine_resize(mni_affine,
mni_shape,
scl,
anchor='f')
if options.prob_mni or options.labels_mni or options.all_mni:
z_mni = spatial.reslice(z, dat_affine, mni_affine, mni_shape)
if options.prob_mni:
fname = options.prob_mni or '{dir}{sep}{base}.prob.mni{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('prob.mni ->', fname)
io.savef(torch.movedim(z_mni, 0, -1),
fname,
like=ref_native,
affine=mni_affine,
dtype='float32')
if options.labels_mni:
fname = options.labels_mni or '{dir}{sep}{base}.labels.mni{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('labels.mni ->', fname)
io.save(z_mni.argmax(0),
fname,
like=ref_native,
affine=mni_affine,
dtype='int16')
del z_mni
if options.bias and (options.bias_mni or options.nobias_mni
or options.all_mni):
bias = spatial.reslice(prm['bias'],
dat_affine,
mni_affine,
mni_shape,
interpolation=3,
prefilter=False,
bound='dct2')
if options.bias_mni or options.all_mni:
fname = options.bias_mni or '{dir}{sep}{base}.bias.mni{ext}'
if len(options.input) == 1:
fname = fname.format(**format_dict)
if options.verbose > 0:
print('bias.mni ->', fname)
io.savef(torch.movedim(bias, 0, -1),
fname,
like=ref_native,
affine=mni_affine,
dtype='float32')
else:
for c, (bias1, ref1) in enumerate(zip(bias, options.input)):
format_dict1 = get_format_dict(ref1, options.output)
fname = fname.format(**format_dict1)
if options.verbose > 0:
print(f'bias.mni.{c+1} ->', fname)
io.savef(bias1,
fname,
like=ref1,
affine=mni_affine,
dtype='float32')
if options.nobias_mni or options.all_mni:
nobias = spatial.reslice(dat, dat_affine, mni_affine, mni_shape)
nobias *= bias
fname = options.bias_mni or '{dir}{sep}{base}.nobias.mni{ext}'
if len(options.input) == 1:
fname = fname.format(**format_dict)
if options.verbose > 0:
print('nobias.mni ->', fname)
io.savef(torch.movedim(nobias, 0, -1),
fname,
like=ref_native,
affine=mni_affine)
else:
for c, (nobias1, ref1) in enumerate(zip(bias, options.input)):
format_dict1 = get_format_dict(ref1, options.output)
fname = fname.format(**format_dict1)
if options.verbose > 0:
print(f'nobias.mni.{c+1} ->', fname)
io.savef(nobias1, fname, like=ref1, affine=mni_affine)
del nobias
del bias
need_iwarp = (options.warp_mni or options.prob_wrp or options.labels_wrp
or options.bias_wrp or options.nobias_wrp or options.all_mni
or options.all_wrp)
need_iwarp = need_iwarp and options.warp
if not need_iwarp:
return
iwarp = spatial.grid_inv(prm['warp'], type='disp')
iwarp = iwarp.movedim(-1, 0)
iwarp = spatial.reslice(iwarp,
dat_affine,
mni_affine,
mni_shape,
interpolation=2,
bound='dft',
extrapolate=True)
iwarp = iwarp.movedim(0, -1)
iaff = mni_affine.inverse() @ dat_affine
iwarp = linalg.matvec(iaff[:3, :3], iwarp)
if (options.warp_mni or options.all_mni) and options.warp:
fname = options.warp_mni or '{dir}{sep}{base}.warp.mni{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('warp.mni ->', fname)
io.savef(iwarp,
fname,
like=ref_native,
affine=mni_affine,
dtype='float32')
# --- Warped space -------------------------------------------------
iwarp = spatial.add_identity_grid_(iwarp)
iwarp = spatial.affine_matvec(dat_affine.inverse() @ mni_affine, iwarp)
if options.prob_wrp or options.labels_wrp or options.all_wrp:
z_mni = spatial.grid_pull(z, iwarp)
if options.prob_mni or options.all_wrp:
fname = options.prob_mni or '{dir}{sep}{base}.prob.wrp{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('prob.wrp ->', fname)
io.savef(torch.movedim(z_mni, 0, -1),
fname,
like=ref_native,
affine=mni_affine,
dtype='float32')
if options.labels_mni or options.all_wrp:
fname = options.labels_mni or '{dir}{sep}{base}.labels.wrp{ext}'
fname = fname.format(**format_dict)
if options.verbose > 0:
print('labels.wrp ->', fname)
io.save(z_mni.argmax(0),
fname,
like=ref_native,
affine=mni_affine,
dtype='int16')
del z_mni
if options.bias and (options.bias_wrp or options.nobias_wrp
or options.all_wrp):
bias = spatial.grid_pull(prm['bias'],
iwarp,
interpolation=3,
prefilter=False,
bound='dct2')
if options.bias_wrp or options.all_wrp:
fname = options.bias_wrp or '{dir}{sep}{base}.bias.wrp{ext}'
if len(options.input) == 1:
fname = fname.format(**format_dict)
if options.verbose > 0:
print('bias.wrp ->', fname)
io.savef(torch.movedim(bias, 0, -1),
fname,
like=ref_native,
affine=mni_affine,
dtype='float32')
else:
for c, (bias1, ref1) in enumerate(zip(bias, options.input)):
format_dict1 = get_format_dict(ref1, options.output)
fname = fname.format(**format_dict1)
if options.verbose > 0:
print(f'bias.wrp.{c+1} ->', fname)
io.savef(bias1,
fname,
like=ref1,
affine=mni_affine,
dtype='float32')
if options.nobias_wrp or options.all_wrp:
nobias = spatial.grid_pull(dat, iwarp)
nobias *= bias
fname = options.nobias_wrp or '{dir}{sep}{base}.nobias.wrp{ext}'
if len(options.input) == 1:
fname = fname.format(**format_dict)
if options.verbose > 0:
print('nobias.wrp ->', fname)
io.savef(torch.movedim(nobias, 0, -1),
fname,
like=ref_native,
affine=mni_affine)
else:
for c, (nobias1, ref1) in enumerate(zip(bias, options.input)):
format_dict1 = get_format_dict(ref1, options.output)
fname = fname.format(**format_dict1)
if options.verbose > 0:
print(f'nobias.wrp.{c+1} ->', fname)
io.savef(nobias1, fname, like=ref1, affine=mni_affine)
del nobias
del bias