Hi Colin,
Are you talking about qboot? This performs inference on diffusion ODFs via residual bootstrap, but there is no spherical deconvolution implemented, at least for now.
Also, there is no deterministic tractography alternative, such as TrackVis, I am not sure where you noticed that?
Qboot can be used to obtain distributions of orientations, which can then be fed to probtrackX to do probabilistic tracking.
Hope this is now clearer,
Stam
----- Original Message -----
From: Colin Reveley
To: [log in to unmask]
Sent: Monday, November 28, 2011 11:40 PM
Subject: [FSL] deterministic tracts, SHD, visualization
I note that there's now an SHD backend to FDT in addition to the bayes method.
and I note that deterministic tractography is mentioned as an option.
and yet, it's not clear what program performs deterministic tractography, or what program might be used to visualise it.
my preference would be something like trackvis. the primary purpose is to explore the data interactively and produce masks for probtrackX. I don't know if I could get trackvis to work with the data. I might well. If anyone has, I would appreciate pointers. Or some other similar system.
I must say I'm grateful for the inclusion of an SHD alternative to bedpostX; I am a fan of bedpostX but I think there is value in comparing probtrackX results from both methods, and as I say in using deterministic tractography for a nice 3d interactive way of making nice masks to hone in on tracts running through many busy gyri.
any help appreciated.
Colin,@Sussex
On 28 November 2011 00:01, FSL automatic digest system <[log in to unmask]> wrote:
There are 3 messages totaling 498 lines in this issue.
Topics of the day:
1. data smoothness and corrections....
2. Determining image acquisition plane (2)
----------------------------------------------------------------------
Date: Sun, 27 Nov 2011 19:48:03 +0000
From: Thomas Nichols <[log in to unmask]>
Subject: Re: data smoothness and corrections....
Dear Torsten,
Randomise doesn't give you the critical cluster size threshold by default,
but it's easy to obtain. If you add the -N option to randomise, for each
corrp image you'll get a .txt file that gives the permutation distribution
of the maximum statistic (whether that's max voxel T, max cluster size, max
TFCE score, whatever). If you find the 95%ile of that distribution, that's
the 5% critical threshold based on permutation. Loading that in Matlab and
finding this percentile is easy:
MaxC=load('permdist.txt');
Nperm=length(MaxC);
sMaxC=sort(MaxC);
Level=0.05;
CritC=sMaxC(ceil(Nperm*Crit))
but you're right, it's something that I always report in papers. I'll try
to get that printed out with, say, the -v flag, in an future version.
My problem with 3dcluster, alphasim, and any Monte Carlo based cluster
inference tool, is that you have to believe in the Gaussian autocorrelation
*and* stationarity that the simulations are based on. VBM data are widely
acknowledged to exhibit nonstationary smoothness, but whenever I've looked
at a FWHM image from FMRI data I see hints of structure there too.
Randomise or any permutation-based procedure will automatically account
for any nonstationarity in the data, and is not vulnerable to errors in
estimated FWHM smoothness (even if the data were stationary).
Permutation is "exact", in that it guaranteed to control false positive
risk with very weak assumptions, but it's not perfect: Parametric models
can provide better power *when* all the assumptions are satisfied [1].
But if lots and lots of people find better results with the Monte Carlo
method than with permutation, it might be that the Monte Carlo method is
inflating significances. The traditional way of comparing methods, with
Monte Carlo simulations of homogeneous smooth Gaussian noise, won't reveal
this (as the parametric assumptions *define* the Monte Carlo method, and
permutation can't out-perform that). A large body of null data with real
(i.e. messy) spatial structured noise would be needed to tested to see if
there is a substantial statistical inefficiency in permutation cluster size
inference.
Hope this helps!
-Tom
[1] However, in all the standard settings, e.g. t-tests, permutation tests
have asymtotic relative efficiency of 1, i.e. will be as powerful as
parametric tests when larger and larger sample sizes are considered.
[2] Random Field Theory makes these assumptions too, and additionally
approximations in the P-value formulas, but these are just more reasons not
to use RFT---thought RFT does at least have a way of handling
nonstationarity.
On Fri, Nov 25, 2011 at 10:02 AM, Torsten Ruest <[log in to unmask]> wrote:
> Hi Steve,
>
> Originally we wanted to get a minimum cluster size as a kind of threshold
> for the feat group level zstat images (obtained from group comparisons
> based on seed voxel analyses on resting state fMRI) as the thresh_zstat
> images appear quite stringently thresholded - that's why 3dcluster appears
> attractive (I hope that's a sound reasoning, I am not a statistician...)
>
> Randomise won't give a minimum cluster size (although the TFCE clusters
> have been tested against a critical cluster size as far as I understand,
> can I actually fetch this critical size from somewhere?), though the output
> will be corrected.
>
> So taking the copes is fine?
>
> Thanks!
>
> Torsten
>
--
__________________________________________________________
Thomas Nichols, PhD
Principal Research Fellow, Head of Neuroimaging Statistics
Department of Statistics & Warwick Manufacturing Group
University of Warwick, Coventry CV4 7AL, United Kingdom
Web: http://go.warwick.ac.uk/tenichols
Email: [log in to unmask]
Phone, Stats: +44 24761 51086, WMG: +44 24761 50752
Fax: +44 24 7652 4532
------------------------------
Date: Sun, 27 Nov 2011 21:43:06 +0000
From: FSL help listserv <[log in to unmask]>
Subject: Determining image acquisition plane
Hi, I have some legacy structural and functional images, and I'm trying to determine whether they were acquired axially, saggitally or coronally. I'm assuming I can determine this using the orientation information from an image (via fslhd), maybe via the qform or sform values, but I'm unsure. Any advice would be very helpful.
Thanks, Chris
Example header -
sizeof_hdr 348
data_type INT16
dim0 3
dim1 128
dim2 128
dim3 21
dim4 1
dim5 1
dim6 1
dim7 1
vox_units mm
time_units s
datatype 4
nbyper 2
bitpix 16
pixdim0 0.0000000000
pixdim1 2.1484375000
pixdim2 2.1484375000
pixdim3 7.0000000000
pixdim4 0.0000000000
pixdim5 0.0000000000
pixdim6 0.0000000000
pixdim7 0.0000000000
vox_offset 0
cal_max 0.0000
cal_min 0.0000
scl_slope 1.000000
scl_inter 0.000000
phase_dim 0
freq_dim 0
slice_dim 0
slice_name Unknown
slice_code 0
slice_start 0
slice_end 0
slice_duration 0.000000
time_offset 0.000000
intent Unknown
intent_code 0
intent_name
intent_p1 0.000000
intent_p2 0.000000
intent_p3 0.000000
qform_name Scanner Anat
qform_code 1
qto_xyz:1 -2.148235 -0.000002 -0.096167 137.485245
qto_xyz:2 -0.011165 1.988851 2.647221 -131.234177
qto_xyz:3 -0.027322 -0.812561 6.479427 -25.610464
qto_xyz:4 0.000000 0.000000 0.000000 1.000000
qform_xorient Right-to-Left
qform_yorient Posterior-to-Anterior
qform_zorient Inferior-to-Superior
sform_name Aligned Anat
sform_code 2
sto_xyz:1 -2.148430 0.005841 0.000736 136.501404
sto_xyz:2 0.005483 1.984913 2.678599 -132.203308
sto_xyz:3 -0.002026 -0.822112 6.467233 -26.192535
sto_xyz:4 0.000000 0.000000 0.000000 1.000000
sform_xorient Right-to-Left
sform_yorient Posterior-to-Anterior
sform_zorient Inferior-to-Superior
file_type NIFTI-1
file_code 2
descrip 1.5T 2D SE\EP TR=15750ms/TE=88ms/FA=90deg 20-Apr-2010 14:10:58.702 Mosaic
------------------------------
Date: Sun, 27 Nov 2011 16:55:32 -0500
From: David Gutman <[log in to unmask]>
Subject: Re: Determining image acquisition plane
There may be a fancier way--- but looknig at the data you have a 128x128
array in the x and y dimensions... and 21 "z" slices... which would
strongly suggest this is acquired axially... it looks like your data is
128x128x21 voxels with a resolution of
2.14x2.14 x 7 mm...
On Sun, Nov 27, 2011 at 4:43 PM, FSL help listserv <[log in to unmask]> wrote:
> Hi, I have some legacy structural and functional images, and I'm trying to
> determine whether they were acquired axially, saggitally or coronally. I'm
> assuming I can determine this using the orientation information from an
> image (via fslhd), maybe via the qform or sform values, but I'm unsure.
> Any advice would be very helpful.
>
> Thanks, Chris
>
>
> Example header -
>
> sizeof_hdr 348
> data_type INT16
> dim0 3
> dim1 128
> dim2 128
> dim3 21
> dim4 1
> dim5 1
> dim6 1
> dim7 1
> vox_units mm
> time_units s
> datatype 4
> nbyper 2
> bitpix 16
> pixdim0 0.0000000000
> pixdim1 2.1484375000
> pixdim2 2.1484375000
> pixdim3 7.0000000000
> pixdim4 0.0000000000
> pixdim5 0.0000000000
> pixdim6 0.0000000000
> pixdim7 0.0000000000
> vox_offset 0
> cal_max 0.0000
> cal_min 0.0000
> scl_slope 1.000000
> scl_inter 0.000000
> phase_dim 0
> freq_dim 0
> slice_dim 0
> slice_name Unknown
> slice_code 0
> slice_start 0
> slice_end 0
> slice_duration 0.000000
> time_offset 0.000000
> intent Unknown
> intent_code 0
> intent_name
> intent_p1 0.000000
> intent_p2 0.000000
> intent_p3 0.000000
> qform_name Scanner Anat
> qform_code 1
> qto_xyz:1 -2.148235 -0.000002 -0.096167 137.485245
> qto_xyz:2 -0.011165 1.988851 2.647221 -131.234177
> qto_xyz:3 -0.027322 -0.812561 6.479427 -25.610464
> qto_xyz:4 0.000000 0.000000 0.000000 1.000000
> qform_xorient Right-to-Left
> qform_yorient Posterior-to-Anterior
> qform_zorient Inferior-to-Superior
> sform_name Aligned Anat
> sform_code 2
> sto_xyz:1 -2.148430 0.005841 0.000736 136.501404
> sto_xyz:2 0.005483 1.984913 2.678599 -132.203308
> sto_xyz:3 -0.002026 -0.822112 6.467233 -26.192535
> sto_xyz:4 0.000000 0.000000 0.000000 1.000000
> sform_xorient Right-to-Left
> sform_yorient Posterior-to-Anterior
> sform_zorient Inferior-to-Superior
> file_type NIFTI-1
> file_code 2
> descrip 1.5T 2D SE\EP TR=15750ms/TE=88ms/FA=90deg 20-Apr-2010
> 14:10:58.702 Mosaic
>
--
David A Gutman, M.D. Ph.D.
Assistant Professor of Biomedical Informatics
Senior Research Scientist, Center for Comprehensive Informatics
Emory University School of Medicine
------------------------------
End of FSL Digest - 26 Nov 2011 to 27 Nov 2011 (#2011-327)
**********************************************************
