*Spm2*

#
All 14 entries tagged

No other Warwick Blogs use the tag *Spm2* on entries | View entries tagged *Spm2* at Technorati | There are no images tagged *Spm2* on this blog

## July 31, 2012

### SPM plot units

## What are the units of a plot in SPM?

In SPM, when you click on “Plot” and ask for a “Contrast estimates and 90% CI”, what exactly are the units of the thing plotted? I get this question so many times I wanted to record my answer once and for all.

The short answer is: 2-3 times an approximation to percent BOLD change, if you have a long-block design; otherwise, it depends. Here’s the long answer.

The final units of a plotted contrast of parameter estimates depends on three things:- The scaling of the data.
- The scaling of the predictor(s) that are involved in the selected contrast.
- The scaling of the selected contrast.

### The scaling of the data

For a first level fMRI analysis in SPM the scaling is done “behind the scenes”... there are no options to change the default behavior: The global mean intensity for each image is computed, and the average of those values (the “grand grand mean”) is scaled to be equal to be 100. However, SPM’s spm_global produces a rather dysmal estimate of typical intracerebral intensity (note that the author is listed as “Anonymous”; to be fair, it worked well on the tightly-cropped, standard-space PET images that it was originally designed for). The estimate of spm_global is generally too low, and so when the data are divided by it and multiplied by 100, the typical brain intensity is well over 100. In my experience, if you examine a beta_XXXX.img corresponding to a constant in a first-level regression model, typically gray matter values range from 200 to 300 when it should be 100; for first-level analyses done in subject-space (where there are even more air voxels) the bias can be even greater. Let’s call the typical gray matter intensity of your BOLD T2* images to be *B* for later correcting this effect (e.g. *B* = 250 or 300 or whatever you find).

At the second level, you depend on the contrast estimates passed up from the first level to have a valid interpretation. So all three of these rules should apply to the first **and** second level models, but the data scaling is done at the first level (no further scaling should be done at the second level; e.g., in the Batch Editor, second-level models have “Global normalisation -> Overall grand mean scaling” set to “No”).

### The scaling of the predictor(s) that are involved in the selected contrast

At the first level, the predictors must be scaled appropriately to give the regression coefficients the same units as the data. For BOLD fMRI, this means the baseline to peak magnitude should be 1.0… i.e. a one unit change in a beta coefficient will result in a one unit change in predicted BOLD effect. In SPM, for *loooong* blocks/epochs, 20 seconds or longer, the scaling is as expected, with the baseline-to-plateau difference being 1.0. For shorter blocks however, the peak can be higher – as this corresponds to the ‘initial peak’ part of the HRF, as the duration is too short for the flat plateau to be established (this ‘early peak’ is due to symmetry of convolution, as it must exactly mirror the post-stimulus overshoot; there is some physiological basis for this as well, though) – or they can be lower. For events, the HRF is normalized to sum to unity, which doesn’t help you predict the peak.

If you have anything but long blocks, then, you have to examine the baseline-to-peak (or if you prefer, baseline-to-trough) height, and make a note of it for later scaling (call it *H*). (The design matrix is in `SPM.xX.X`

, in `SPM.mat`

; plot individual columns to judge the scaling). For event-related designs, however, it can be tricky to uniquely define the baseline-to-peak height; if you have a slow design, with isolated events, you’ll find a different peak height than if you have a design with some closely-spaced events as the responses pile-up additively. A reasonable unique definition of a event-related response magnitude is the height of an isolated event; again, call this *H*.

At the second level, you must also make sure that your regressors preserve the units of the data. For example, if you are manually coding dummy regressors, like gender, as -1/1, realize that the unit effect of beta coefficient for gender will result in a **two** unit change in the data. Safer is using a 0/1 coding, though that then re-defines the intercept; best yet is to use -1/2 and 1/2 as dummy values for a dichotomous variable.

For parametric modulation and covariates, note that a regression coefficient’s units are determined by the units of the covariate. For example, if you have a parametric modulation in a reward experiment, where the units are £’s that are at risk in a bet, then your regressor coefficient has units (if everything else is working as it should) “percent BOLD change per £ at risk”. Likewise, if you use an age predictor in the second level, if you enter age in years (centered or uncentered), the units of the age coefficient are “percent BOLD change per year”

Finally, beware of trying to compare between percent BOLD change in Block Designs vs. Event Related designs. Due to the nonlinearities of the BOLD effect, a percent BOLD change for an event related response is really a different beast from a percent BOLD change for an block design. When reporting one type, be sure it is clear what sort of effect you’re reporting, and don’t try averaging or comparing their magnitudes.

### The scaling of the selected contrast

The General Linear Model is a wondrous thing, as you’ll get the same T-values and P-values for the contrast [-1 -1 1 1] as you do for [-1/2 -1/2 1/2 1/2]. However, the interpretation of the contrast of parameter estimates (contrast times beta-hats) is different for each. The following rules to scale contrasts will*usually*work to ensure that a contrast preserves the units of the regression coefficients:

- Sum of any positive contrast elements is 1.
- Sum of any negative contrast elements is -1.

If the contrast elements are all positive or all negative, the interpretation is clear, as you are ensuring that the beta-coefficents (each which should have unit-change interpretations) will be averaged, giving a (average) unit change interpretation. If there are a mix of positive and negative elements, I see this rule as carefully ensuring there is a positive unit change quantity subtracted from a negative unit change quantity, giving the expected difference of unit change. (This rule should work for covariates and simple 0/1 dummy variables; this FSL example shows a case with -1/0/1 dummy variables where the contrasts violate these rules yet actually give the correct units.)

### When it goes right, when it goes wrong

*If* everything had gone well, then the plotted units are “approximate percent BOLD change”. The “approximate” qualification results from only using a global-brain mean of 100; an “exact” percent BOLD change requires that the voxel in question has had baseline (or grand mean) intensity of exactly 100. We don’t grand-mean-normalise voxel-by-voxel because of edge effects: At the edge of the brain, the image intensity falls off to zero, and dividing by these values will give artifactually large values.

Of the three scalings in play, the first two of these three items are basically out of the control of the user. As hinted, first-level fMRI data in SPM are generally mis-scaled, resulting in global brain intensity between 200 and 300, i.e., 2 to 3 times greater than anticipated. Also, if you don’t have long blocks, you’ll need to apply a correction for the non-unity *H* you measured. The corrections are as follows: Dividing the reported results *B*/100 (i.e. multiplying by 100/*B*) will correct the problem of the data being incorrectly scaled. Multiplying by *H* will correct the units of the regression coefficients. After both of these corrections, you should get something closer to approximate percent BOLD change.

This all sounds rather depressing, doesn’t it? Well, there is a better way. If you use the Marsbar toolbox it will carefully estimate baseline intensity for the ROI (or single-voxel ROI—make sure it isn’t on the edge of the brain though!) and gives exact percent BOLD change units.

### Postscript: What about FSL?

FSL has an good estimate of the global mean, because it uses morphological operations to identify intracerebral voxels. It scales grand mean brain intensity to 10,000, however (because it used to use a integer data representation and 100 would have resulted in too few significant digits.) Block-design predictors are scaled to have unit height; the default HRF doesn’t use a post-stimulus undershoot, and so baseline-to-peak height is the same as trough-to-peak height. Event-related predictors *are* *not* scaled to have peak height of unity; however, the recommended tool for extracting percent BOLD change values, `featquery`

, estimates a trough-to-peak height and then appropriate scales the data. However, as Jeanette Mumford has pointed out, this isn’t really the right thing to do when you have closely spaced trials; see her excellent A Guide to Calculating Percent Change with Featquery to see how to deal with this.

#### Other references

- Percent Signal Change for fMRI calculations by Paul Mazaika.

- How is the percent signal change calculated? from the MarsBar FAQ.

- Percent Signal Change FAQ from the MIT Mindhive on brain research.

- Scaling and Percent Signal Change from AFNI’s Gang Chen.

*Thanks to Jeanette Mumford & Tor Wager for helpful input and additions to this entry, and Guillaume Flandin for the additional references.*

## June 03, 2010

### SPM2 Gem 13: Corrected cluster size threshold

This is a script that will tell you the corrected cluster size threshold for given cluster-defining threshold: corrclusth.m

The usage is pretty self explanatory:

Find the corrected cluster size threshold for a given alpha function [k,Pc] =CorrClusTh(SPM,u,alpha,guess) SPM - SPM data structure u - Cluster defining threshold If less than zero, u is taken to be uncorrected P-value alpha - FWE-corrected level (defaults to 0.05) guess - Set to NaN to use a Newton-Rhapson search (default) Or provide a explicit list (e.g. 1:1000) of cluster sizes to search over. If guess is a (non-NaN) scalar nothing happens, except the the corrected P-value of guess is printed. Finds the corrected cluster size (spatial extent) threshold for a given cluster defining threshold u and FWE-corrected level alpha.

To find the 0.05 (default alpha) corrected cluster size threshold for a 0.01 cluster-defining threshold:

>> load SPM >> CorrClusTh(SPM,0.01) For a cluster-defining threshold of 2.4671 the level 0.05 corrected cluster size threshold is 7860 and has size (corrected P-value) 0.0499926

Notice that, due to the discreteness of cluster sizes you cannot get an exact 0.05 threshold.

The function uses an automated search which may sometimes fail. If you specify a 4th argument you can manually specify the cluster sizes to search over:

>> CorrClusTh(SPM,0.01,0.05,6000:7000) WARNING: Within the range of cluster sizes searched (6000...7000) a corrected P-value <= alpha was not found (smallest P: 0.0819589) Try increasing the range or an automatic search.

Ooops... bad range.

Lastly, you can use it as a look up for a specific cluster size threshold. For example, how much over the 0.05 level would a cluster size of 7859 be?

>> CorrClusTh(SPM,0.01,0.05,7859) For a cluster-defining threshold of 2.4671 a cluster size threshold of 7859 has corrected P-value 0.050021

Just a pinch!

### SPM2 Gem 12: fMRI Analysis threshold

This is an update of Gem12 for SPM99, originally by Stefan Kiebel.

> Is it possible to instruct spm99 to search all voxels within a given > mask image rather than all above a fixed or a %mean threshold? Yes, with SPM2 it's possible to use several masking options. To recap, there are 3 sorts of masks used in SPM2: 1. an analysis threshold 2. implicit masking 3. explicit masking 1: One can set this threshold for each image to -Inf to switch off this threshold. 2: If the image allows this, NaN at a voxel position masks this voxel from the statistics, otherwise the mask value is zero (and the user can choose, whether implicit masking should be used at all). 3: Use mask image file(s), where NaN (when image format allows this) or a non-positive value masks a voxel. On top of this, SPM automatically removes any voxels with constant values over time. So what you want is an analysis, where one only applies an explicit mask. In SPM2 for PET, you can do this by going for the Full Monty and choosing -Inf for the implicit mask and no 0-thresholding. Specify one or more mask images. (You could also define a new model structure, controlling the way SPM for PET asks questions). With fMRI data/models, SPM2 is fully capable of doing explicit masking, but the user interface for fMRI doesn't ask for it. One way to do this type of masking anyway is to change the SPM.mat file *after* you specify your model, but *before* clicking 'Estimate'. Specifically: 1. Load the SPM.mat file, load SPM set the SPM.xM.TH values all to -Inf, SPM.xM.TH = -Inf*SPM.xM.TH; and, in case that you have an image format not allowing NaNs, set SPM.xM.I to 0 SPM.xM.I = 0; 2. If using a mask image, set SPM.xM.VM to a vector of structures, where each structure element is the output of spm_vol. For instance: SPM.xM.VM = spm_vol('Maskimage'); 3. Finally, save by save SPM SPM

### SPM2 Gem 11: Un–normalization

This script of John's will find the corresponding co-ordinate in un-normalised image: get_orig_coord2.m

### SPM2 Gem 10: Creating a mask from a XYZ list

This is a "Jan's Gem" from Jan Gläscher...

Subject: Re: [SPM] creating a mask from voxel coordinates From: Jan Gläscher <glaescher@UKE.UNI-HAMBURG.DE> Date: Tue, 15 Mar 2005 12:24:27 +0100 To: SPM@JISCMAIL.AC.UK Dear Susana, you can try the following recipe. It might not be the easiest way, but it should work. Suppose your n MNI coordinates are saved in a 3xn matrix called mni. (This is essential (not the name, but the format of 3xn), otherwise the steps below won't work). 1. Create an spm_vol handle to the image, that you determined the coordinates from Vin = spm_vol(spm_get(1,'*.img','Select image')); 2. Read the information in the image. (The data are not needed, this is done purely for the purpose of setting up a matrix of voxel coordinates. [Y,XYZ] = spm_read_vols(Vin); 3. Setup a matrix of zeros in the dimensions of the input image mask = zeros(Vin.dim(1:3)); 4. Now loop over all voxels in your mni variable and set the corresponding location in the mask matrix to 1: for v = 1:size(mni,2) mask(find(XYZ(1,:)==mni(1,v)&XYZ(2,:)==mni(2,v)&XYZ(3,:)==mni(3,v))) = 1; end 5. Setup an spm_vol output file handle and change the filename Vout = Vin; Vout.fname = '/path/to/the/directory/mask.img'; 6. Finally, write the new mask to the output file. spm_write_vol(Vout,mask); Cheers, Jan

### SPM2 Gem 9: Masking w/ ImCalc

This is an email from Rik Henson which references John. It addresses the fact that ImCalc defaults to writing out images in 2-byte signed integer format, which is can be problematic if you're masking statistic images (which should written with floating point precision).

Subject: Masking with ImCalc From: Rik Henson <r.henson@UCL.AC.UK> Date: Fri, 14 May 2004 08:24:25 +0100 (03:24 EDT) To: SPM@JISCMAIL.AC.UK Several people have asked how to mask SPMs across different analyses. This used to be possible using ImCalc in SPM99 (see Mailbase archive). In SPM2 however, when ImCalc is called from the GUI, it adopts various defaults, which include writing images as "uint16" (rather than "float", which is the normal datatype for T/F imgs). Thanks to John for pointing this out. It is for this reason that the procedure of overwriting spmT/F*imgs using ImCalc does not work from SPM2's GUI. To override this default however, you can simply call ImCalc from the command line instead, setting the datatype to spm_type 'float': spm_imcalc_ui([],[],'i1.*(i2>3.09)',{[],[],'float',0}) The empty 1st and 2nd arguments mean that SPM will prompt you to select the input images and the output filename, as usual. You will of course need to change the third argument to your particular ImCalc equation for evaluation. The 4th argument is a cell array of defaults, which includes the 'float' enforcement. For more info, type "help spm_imcalc_ui". Rik

See also SPM2 Gem 8, and SPM99 Gems 3, 16, and 21.

### SPM2 Gem 8: ImCalc Script

SPM99 Gem 21 is the same (with minor changes).

Subject: Re: a script to use ImaCal From: John Ashburner <john@FIL.ION.UCL.AC.UK> Date: Thu, 16 Oct 2003 11:24:17 +0000 (07:24 EDT) To: SPM@JISCMAIL.AC.UK > Brain image for each subject to mask out CSF signal was generated by > using MPR_seg1.img (i1) and MPR_seg2.img (i2) with (i1+i2)>0.5 in > ImaCal.(called brainmpr.img for each subject) > > Then I have more than two-hundred maps, which need to mask out > CSF. I think I can use ImaCal again with selecting brainmpr.img (i1) > and FAmap.img (i1), and then calculating (i1.*i2) to generate a new > image named as bFAmap.img. Unfortunately, if I use the ImaCal, it > take so long time to finish all subjects. Could anyone have a script > to generate a multiplication imaging with choosing a brain > image(brainmpr.img, i1) and a map image (FAmap.img, i2) and writing > an output image (bFAmap.img) from i1.*i2? You can do this with a script in Matlab. Something along the lines of the following should do it: P1=spm_get(Inf,'*.IMAGE','Select i1'); P2=spm_get(size(P1,1),'*.IMAGE','Select i2'); for i=1:size(P1,1), P = strvcat(P1(i,:),P2(i,:))); Q = ['brainmpr_' num2str(i) '.img']; f = '(i1+i2)>0.5'; flags = {[],[],[],[]}; Q = spm_imcalc_ui(P,Q,f,flags); end; Note that I have not tested the above script. I'm sure you can fix it if it doesn't work. Best regards, -John

### SPM2 Gem 7: Extracting realignment parameters

The referenced script is here save_parameters.m

Subject: Re: [SPM] realignment parameter From: John Ashburner <john@FIL.ION.UCL.AC.UK> Date: Fri, 23 Apr 2004 11:25:27 +0000 (07:25 EDT) To: SPM@JISCMAIL.AC.UK > I have switched from spm99 to spm2 and am learning new things. > When I realign images interactively with GUI interface, > the rp*.txt file is created. However, the parameter file > is not created - or, I cannot find it -, when I realign > with a batch script. > I am sure I left out something, but I don't know it. > Any suggestion will be very appreciated. If you call spm_realign with output (left-hand) arguments, then no rp_*.txt file is saved. You could create a save_parameters.m file, containing: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% function save_parameters(V) fname = [spm_str_manip(prepend(V(1).fname,'rp_'),'s') '.txt']; n = length(V); Q = zeros(n,6); for j=1:n, qq = spm_imatrix(V(j).mat/V(1).mat); Q(j,:) = qq(1:6); end; save(fname,'Q','-ascii'); return; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% and derive the values from your realigned images: P = spm_get(Inf,'*.IMAGE','Select realigned images'); save_parameters(spm_vol(P)) Best regards, -John

### SPM2 Gem 6: Contrast SE Computation

An email from Will Penny on how a contrast's standard error is computed.

Subject: Re: Beta SE in SPM2? From: Will Penny <wpenny@FIL.ION.UCL.AC.UK> Date: Fri, 27 Feb 2004 11:26:35 +0000 To: SPM@JISCMAIL.AC.UK Tobias, > How do I extract beta SE in SPM2? As far as I can see I can only get > y, Y, beta, and Bcov...have I missed something? If you're referring to the Standard Error or the Standard deviation of the parameter estimates then this isn't stored explicitly. Its computed on the fly when one makes an inference about a contrast. See the following lines (153-158) in spm_contrasts: case 'T' %-Compute SPM{t} image %--------------------------------------------------------------- cB = spm_get_data(xCon(ic).Vcon,XYZ); l = spm_get_data(VHp,XYZ); VcB = xCon(ic).c'*SPM.xX.Bcov*xCon(ic).c; Z = cB./sqrt(l*VcB); So, if you did a contrast xCon(ic).c'=[1 0 0 0] to test the null hypothesis that the first parameter is zero then cB will be equal to beta1 and sqrt(l*VcB) will be equal to std(beta1) - the standard deviation of the first parameter estimate. Note that l is the residual sum of squares at that voxel which is stored in the image ResMS.img and accessed via the file handle VHp. In this way SPM does'nt have to store Standard Deviations/covariances for every voxel and beta - so saving disk space. But its not that handy for what you want ! Hope this helps, Will.

### SPM2 Gem 5: Normalizing [x y z]

An off-list email reports how to find the point in an normalized image that corresponds to a given un-normalized location. See also Un-normalizing a point.

From: John Ashburner <john@fil.ion.ucl.ac.uk> To: Christophe Phillips <c.phillips@ulg.ac.be> Subject: Re: Coordinates transform Date: Wed, 4 Feb 2004 17:26:21 +0000 Hi Christophe, > I've the coordinates [x y z] of a point (actually the location of a > fitted current dipole) in an image 'img1'. This image was normalised > into 'w_img1'. I'd like to have the dipoles coordinates [x y z]_w in > the normalised image... > > Do you have a routine that can do that easily from the 'img1_sn.mat' > file ? The deformations in the sn.mat files map from points in normalised space, to points in un-normalised space. You therefore need to invert them, which is a bit messy. You would use the Deformations toolbox for this (from the Toolbox pulldown). First write out the deformation field from the sn.mat. Then call the Deformations toolbox again, and choose the deformations inversion option. This will be an iy*.img file with three volumes. This little script can then be used. c = [x y z]'; % co-ordinates in mm (as shown by Display) P = spm_get(1,'iy*.img','Select inverted warp'); P = [repmat(P,3,1) [',1';',2';',3']]; V = spm_vol(P); vx = inv(V(1).mat)*[c ; 1]; % The voxel in the deformation to sample w_coord = [... spm_sample_vol(V(1),vx(1),vx(2),vx(3),1) spm_sample_vol(V(2),vx(1),vx(2),vx(3),1) spm_sample_vol(V(3),vx(1),vx(2),vx(3),1)] I haven't tested it, so there may be things to fix. All the best, -John