Markus -
> Dear Rik!
>
> Thank you so much for your detailed comments!
> I am doing a divided visaul field study with stimulus presentation
> times of 200 ms and ISI von jittered ~4500ms (3000 - 6000ms), that
> should be ok for the latency calculation?
>
Well, it won't have very high efficiency, but it might be sufficient...
> Behaviorally I observe an left-right asymmetry in response time. Thats
> why I would be very much interested to see if there is an activity
> crossing the corpus callosum, causing latency, but helping this
> asymmetrical effect happen.
>
> I do not yet understand right how I chould mask the regions to look at
> with the activation of the canonical (as I understand your
> recommendation).
> For example I have a contrast Bilateral_vs_Left, giving me the
> leftovers at right, and other contrast Bilateral_vs_Right giving the
> leftovers at left. I my case the leftovers at left show only a small
> occipital blob, when modeled with the canonical only. Those at right
> are much bigger.
> When modeled with canonical and temp deriv (without interest, i.e. [1
> 0]), this does not change. But when modeling both hrf and td [1 1],
> this small blob at left becomes much bigger extrastriate activation
> (similar to the one at right).
>
Unfortunately a T-contrast [1 1] on hrf and temporal derivative is
fairly meaningless - it is averaging "apples and pears". The canonical
and tds need to be selected in different contrasts; or perhaps easier,
you can construct different 2nd-level models for the canonical and the
td. What I meant about masking is to evaluate con*imgs for the 1st-level
contrast [1 0 1 0] that looks at the average effect of can hrf across
both left and right events, and put these into a one-sample t-test. Then
calculate the 1st-level contrast [0 1 0 -1] that tests for differences
in the canonical, and put this into a one-sample t-test. Then when you
test for a [1] or [-1] T-contrasts in the latter, mask the results
(inclusively, eg through SVC) with the [1] F-contrast for canonical hrf
one-sample t-test (at a reasonably strict threshold). This will restrict
differences in td to those regions where the canonical HRF captures
significant variance.
Note however that you cannot interpret any td differences you find in
terms of latency until you scale them by their respective canonical hrf
estimates (because a large hrf needs a large td to shift it). This
requires calculating the ratio in ImCalc, as in your previous email, or
if you want a rough and ready way, you could also mask your [1] or [-1]
T-contrasts on the td with a [1]-Fcontrast from a third, 2nd-level
one-sample t-test for the 1st-level contrast [1 0 -1 0] - ie to capture
the basic left/right canonical hrf differences you have found - this
mask should be "exlusive", ie, removing voxels where there is a
difference in the can hrf - which means that any differences in the td
are probably true latency differences.
Sorry - this is getting complex I know - but please search the archives
where I think more details about such masking can be found.
> My hypothesis was thus that wen integrating the latencies in a correct
> way, I could show that there was activation on the ipsilateral
> hemisphere, but with a latency that is due to the interhemispheric
> information flow.
> However when I have to mask the td image with the hrf activation (as
> you mention), I will not be able to see this additional activation.
>
> If you think my reasoning is correct, then I would like to proceed
> with the latency calculation! ;-)
> I would also be very grateful if you could send me the mentioned
> script calculating the latency images!
>
Will do.
R
> Thank you so much again!
> Best, Markus
>
>
>
> 2009/11/27 Rik Henson <[log in to unmask]>:
>
>> Markus -
>>
>> First, you might want to check this archived message:
>>
>>
>> https://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0904&L=SPM&D=0&X=1B29E63F8DCB1C3993&Y=rik.henson%40mrc-cbu.cam.ac.uk&P=196039
>>
>> where Donald McLaren identified a problem with using the precise sigmoidal
>> equation in Henson et al (2002) for data analysed in SPM5+, because the
>> scaling in spm_get_bf.m changed (that paper was based on SPM2).
>>
>> To answer your specific questions:
>>
>>
>>> I am refering to the paper of Henson et al 2002 NeuroImage 15, 83-95,
>>> about the Latency differences.
>>>
>>> I would like to use this technique.
>>>
>>> I am right when I proceed as follows?
>>>
>>> a) integrate temporal derivatives into my first level model
>>>
>> Yes. Note however that you can only separate estimation of latency from
>> estimation of height of an HRF if you have long or jittered SOAs (eg null
>> events); with rapid, fixed-SOA event-related designs (eg SOA<~2s, randomised
>> event-types), the temporal derivative for one event-type will be correlated
>> with the difference in (ie contrast of) canonical HRFs across event-types.
>> This is just an example of the more general point that you need to estimate
>> each regressor (temporal basis function) with high statistical efficiency -
>> ie the distinction between estimating an HRF *shape* and simply detecting
>> the *amplitude* of an assumed shape (e.g, relative efficiencies of a
>> canonical HRF vs an FIR basis set; see Henson, 2004, HBF book chapter).
>>
>>
>>
>>> b) calculate the beta for the hrf_<each condition> as a [1 0]
>>>
>>> c) calculate the beta for the td_<each condition> as a [0 1]
>>>
>>> d) use the image calculator to create the latency_image_<each condition>
>>> using the formula
>>>
>>> 2C/(1+exp(D beta2/beta1)) - C;
>>>
>>> where C=1.78, D=3.1 (Henson et al 2002, Neuroimage 15, p86).
>>>
>> Yes to points b-d, except that you might need to restimate the parameters C
>> and D in you are using SPM5+, as Donald found.
>>
>> Note that these parameters only matter if you want to estimate the precise
>> latency (eg in seconds), which is only really valid in the linear regime
>> where the Taylor approximation holds (ie +/-1s of the canonical latency).
>> Furthermore, precise latency differences in the BOLD impulse response may
>> not be easily interpretable, because they do not necessarily reflect latency
>> differences in the underlying neural activity (which is what I assume you
>> are really interested in) - given the time integration (see Discussion in
>> Henson et al, 2002) and that the neural-BOLD coupling is likely to have
>> appreciable nonlinearities. (This is perhaps one reason that the various
>> published methods for estimating BOLD latencies have not been used
>> extensively for neuroscientific conclusions.)
>>
>> If you don't care about precise latency, then you can view the sigmoidal
>> function just as a statistical transform that prevents the
>> derivative:canonical ratio from exploding beyond the linear regime (or when
>> the canonical estimate is close to zero - ie for voxels where there is no
>> basic impulse response in the first place). Then the precise parameters
>> don't matter: you are just conditioning the data so that it becomes more
>> Gaussian (the ratio won't be precisely Gaussian, even after transformation
>> (you could use a log transform for that), though with enough Gaussian
>> smoothing, the parametric stats should be reasonably robust). It also helps
>> to only analyse voxels where there is a significant loading on the canonical
>> HRF as well (ie use an inclusive mask, as in Henson et al, 2002), where the
>> ratio only really makes sense (as mentioned above).
>>
>>
>>
>>> e) enter these latency_images into the second level stats (ANOVA).
>>>
>> Yes
>>
>>
>>
>>> f) And the last question: Is the above formula correctly entered into
>>> the ImageCalculator when doing:
>>>
>>> f = '(2*1.78./(1+exp(3.1*i2./i1))) - 1.78' ( I mean, I do get some
>>> imges, but are they correct??)
>>>
>> Should be. I can send you a function that writes latency images offline if
>> you want. But only if you are sure you want to proceed with latency
>> analyses.... ;-)
>>
>> Rik
>>
>>
>> --
>>
>> -------------------------------------------------------
>> Dr Richard Henson
>> MRC Cognition & Brain Sciences Unit
>> 15 Chaucer Road
>> Cambridge
>> CB2 7EF, UK
>>
>> Office: +44 (0)1223 355 294 x522
>> Mob: +44 (0)794 1377 345
>> Fax: +44 (0)1223 359 062
>>
>> http://www.mrc-cbu.cam.ac.uk/people/rik.henson/personal
>> -------------------------------------------------------
>>
>>
>>
>>
>
>
>
>
--
-------------------------------------------------------
Dr Richard Henson
MRC Cognition & Brain Sciences Unit
15 Chaucer Road
Cambridge
CB2 7EF, UK
Office: +44 (0)1223 355 294 x522
Mob: +44 (0)794 1377 345
Fax: +44 (0)1223 359 062
http://www.mrc-cbu.cam.ac.uk/people/rik.henson/personal
-------------------------------------------------------
