Hi Christina,
On 27 Dec 2006, at 15:13, Christina Hugenschmidt wrote:
> Thanks Christian! Your response was most helpful. I have a few more
> questions about ICA group analysis based on your response,papers, and
> the archived posts I have read. I have run ICA on all my subjects, I
> have identified the component of interest, but I am not sure what
> to do
> next. By way of reference, I am most familiar with GLM analyses.
>
> I would like to compare one component in old vs. young subjects, or if
> possible, treating age as a continuous variable (we have a middle-aged
> cohort as well).
>
> There seem to be a number of different approaches to doing a group
> analysis with ICA. If possible, I would prefer to do a random-effects
> style analysis where both between and within subject variance is
> modeled.
>
The issues with group ICA are largely the same as in standard GLM
group analysis
> 1. Is it best to use the z-maps for a group analysis?
This is currently an easy practical solution. You could feed the ICA
Z-stats maps into a higher-level GLM - this then constitutes a poor-
man's random effects analysis as it assumes identical within subject
variances (in exactly the same way as GLM group analysis used to be
carried out in popular fmri analysis packages ;)
Alternatively, you could also concatenate the Z-stats maps and run
PCA on this.
> 2. What is the most statistically appropriate way to combine
> components
> to make a group map?
My take on this is to directly decompose all data using the tensor
approach - the code is now being tested across different data sets
and next version will make it more widely available as part of the
next release version of FSL.
> 3. You mentioned doing spatial correlations to identify components
> that
> are common across subjects. Could you expand on that a little more?
> Does
> this correlation use the GLM?
>
nope, you simply run 'avwcc 1stfile 2ndfile threshold'
to get a list of pairs of volume indices in the first and the second
file where the spatial correlation exceeds the threshold level. Both
the 1st and 2nd file can be 4D files.
hope this helps
cheers
Christian
> Thank you again,
>
> Christina
>
> Christina Hugenschmidt
> Graduate Student, Neuroscience Program
> Wake Forest University School of Medicine
> [log in to unmask]
> (336) 716-0972
>
>
> -----Original Message-----
> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
> Behalf Of Christian Beckmann
> Sent: Wednesday, December 20, 2006 5:05 PM
> To: [log in to unmask]
> Subject: Re: [FSL] Basic ICA Questions
>
> Hi,
>
> the order of compoennts after a standard decomposition is not
> fundamentally meaningful - unlike a PCA decomposition the time
> courses are not mutually orthogonal and therefore do not explain
> separate amounts of variance.
> Given that there is no natural order it's then sensible to order
> components according to some criterion, e.g. in the current release
> version melodic orders the components in order of decreasing amounts
> of _uniquely_ explained variance. This then is meaningful...
>
> Wrt number of components: yes, we believe that selecting the number
> of components based on an estimate of the amount of non-Gaussianity
> contained in the data is a good thing - picking a number at random
> has various problems, some of which are summarised in the relevant
> technical report.
>
> Wrt group analysis: don't know what you mean by 'group analysis'. The
> current release version does not do group ICA... for identifying
> "corresponding" maps between decompositions. This is difficult but
> can be achieved using e.g. spatial correlations (use avwcc for this).
> The next release version will include multi-subject decompositions,
> including Tensor-ICA, simple concatenation and PARAFAC, so the
> problem of finding correspondences then is solved as part of the
> decomposition... watch this space.
> cheers
> Christian
>
>
> On 20 Dec 2006, at 20:08, Christina Hugenschmidt wrote:
>
>> Hi All,
>>
>> I am interested in trying an ICA analysis of some of my data, and
>> have been doing some reading. As a complete novice, though, I still
>> have some questions. They are rather basic, but I am appreciative
>> of any insights people could offer.
>>
>>
>> Is the order of the components meaningful?
>> In reading the technical papers about MELODIC, it seems that
>> allowing FSL to choose the number of components is a good option.
>> However, in reading the literature, it seems that people usually
>> specify a set number of components. Which is the better option? If
>> you specify a number of components, how do you choose a good
>> number? My understanding from reading is that if a subject does not
>> converge onto the specified number of components, that data is
>> excluded from further group analysis. Is this so?
>> Does anyone have recommendations for how to efficiently identify a
>> desired network, so that that component can be used in a group
>> analysis?
>>
>> Thank you!
>>
>>
>> Christina
>>
>>
>> Christina Hugenschmidt
>>
>> Graduate Student, Neuroscience Program
>>
>> Wake Forest University School of Medicine
>>
>> [log in to unmask]
>>
>> (336) 716-0972
>>
>>
>>
>
> --
> Christian F. Beckmann
> Oxford University Centre for Functional
> Magnetic Resonance Imaging of the Brain,
> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> Email: [log in to unmask] - http://www.fmrib.ox.ac.uk/
> ~beckmann/
> Phone: +44(0)1865 222551 Fax: +44(0)1865 222717
--
Christian F. Beckmann
Oxford University Centre for Functional
Magnetic Resonance Imaging of the Brain,
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
Email: [log in to unmask] - http://www.fmrib.ox.ac.uk/~beckmann/
Phone: +44(0)1865 222551 Fax: +44(0)1865 222717
|