fmri.smooth {fmri} | R Documentation |
Perform the adaptive weights smoothing procedure
fmri.smooth(spm, hmax = 4, adaptive=TRUE, lkern="Triangle", skern="Triangle", na.rm=FALSE)
spm |
object of class fmrispm |
hmax |
maximum bandwidth to smooth |
adaptive |
logical. TRUE (default) for adaptive smoothing |
lkern |
lkern specifies the location kernel. Defaults to
"Triangle", other choices are "Gaussian", "Quadratic", "Cubic" and
"Uniform". Note that the location kernel is applied to
(x-x_j)^2/h^2 , i.e. the use of "Triangle" corresponds to the
Epanechnicov kernel in nonparametric kernel regression. |
skern |
skern specifies the kernel for the statistical
penalty. Defaults to "Triangle", the alternative is "Exp".
lkern="Triangle" allows for much faster computation (saves up
to 50%). |
na.rm |
na.rm specifies how NA's in the SPM are handled. NA's may occur
in voxel where the time series information did not allow for estimating parameters and their variances
or where the time series information where constant over time. A high (1e19) value of the variance
and a parameter of 0 are used to characterize NA's. If na.rm=FALSE these values are simply downweigthed
by their high variance and estimates in these voxels are produced as nonadaptive averages
of values from neighboring voxels. If na.rm=TRUE no estimates are computed (the values
for mean and variance are kept) in these voxels. |
This function performs the smoothing on the Statistical Parametric Map spm.
hmax
is the (maximal) bandwidth used in the last iteration. Choose
adaptive
as FALSE
for non adaptive
smoothing. lkern
can be used for specifying the
localization kernel. For comparison with non adaptive methods use
"Gaussian" (hmax given in FWHM), for better adaptation use "Triangle"
(default, hmax given in voxel). skern
can be used for specifying the
kernel for the statistical penalty.
The function handles zero variances by assigning a large value (1e20) to these variances.
object with class attributes "fmrispm" and "fmridata"
cbeta |
smoothed parameter estimate |
var |
variance of the parameter |
hmax |
maximum bandwidth used |
rxyz |
smoothness in resel space. all directions |
rxyz0 |
smoothness in resel space as would be achieved by a Gaussian filter with the same bandwidth. all directions |
scorr |
spatial correlation of original data |
weights |
ratio of voxel dimensions |
vwghts |
ratio of estimated variances for the stimuli given by
vvector |
hrf |
Expected BOLD response for the specified effect |
Joerg Polzehl polzehl@wias-berlin.de, Karsten Tabelow tabelow@wias-berlin.de
Tabelow, K., Polzehl, J., Voss, H.U., and Spokoiny, V. (2005). Analysing {fMRI} experiments with structure adaptive smoothing procedures, NeuroImage, accepted (2006).
Polzehl, J. and Spokoiny, V. (2006). Propagation-Separation Approach for Local Likelihood Estimation, Probab. Theory Relat. Fields 135, 335-362.
## Not run: fmri.smooth(spm, hmax = 4, lkern = "Gaussian")