Hi Jesper,
thanks for the information.
On Sep 26, 2012, at 14:22 , Jesper Andersson wrote:
> Hi again,
>
>>> There is also the issue of susceptibility dephasing in the gradient echo
>>> that causes unrecoverable signal loss, which is much less present in spin
>>> echo data.
>>
>> Well, isn't that a problem for all gradient echo sequences independent of topup (in other words prelude and fugue will also not recover that signal)?
>
> Yes and no, and this is a crucial point. When you use a "traditional" fieldmap to correct a GE-EPI it is true that it cannot help you to correct the lost signal, but at least it tells you how to "shift" the existing signal (because the fieldmap is defined everywhere).
>
> If you were to use PE-flipped GE-EPI images to calculate the fieldmap (a la topup) it won't even be defined in the areas with no/little signal.
>
> One could always turn that argument around and say that in areas with little no signal there is anyway nothing to shift, but I would always prefer to use the SE-EPI topup derived fieldmaps (as Matt described us doing in the HCP project).
I see, too bad
>
>>>
>>> It is possible to acquire a few phase encoding direction reversed spin
>>> echo scans and calculate the field map from them using topup and then
>>> apply this to the gradient echo data just like any other field map (this
>>> is what we are doing in the HCP). This will not solve any motion related
>>> issues, however.
>>
>> I see, trying to interleave spin echo pairs into an gradient echo sequence (to sort of track the motion induced susceptivity changes) does not seem to be feasible… Since the apparent motion is one of mu biggest issues I would settle for fixing that first. (To go all overboard, maybe it is possible to combine both approaches, use topup to get rid of the apparent motion and get a more homogenous native functional space, reapply the canonical warp field and then use a spin-echo derived field map to recover the lost signal (and get rid of the warping). My gut feeling tells me that it probably is impossible)
>
> Two issues here
>
> 1. Tracking the changing field
> 2. Recovering the lost signal
>
> 1. It would be possible to track the changing field using a blip-reversed pair for each time point, but that would double your TR. If the changing field is really a problem for you I would instead suggest using the method in Andersson et al. NeuroImage 2001.
I guess its reading time, thanks for the pointer, do yu whether anyone has implemented that method already?
>
> 2. This is unfortunately never going to happen. If the signal was never recorded there is no getting it back (in a meaningful way).
Well, it would have been nice….
Best
Sebastian
>
> Good Luck Jesper
>
>
>
>>
>> best
>> Sebastian
>>
>>>
>>> Peace,
>>>
>>> Matt.
>>>
>>> On 9/26/12 1:36 PM, "Sebastian Moeller"
>>> <[log in to unmask]> wrote:
>>>
>>>> Hi Jesper,
>>>>
>>>> I am quite impressed, many thanks for creating and supplying quality
>>>> analysis tools (for the rest of usŠ).
>>>> I have one quick question on topup; will it work with standard bold EPIs
>>>> as well?. The wiki talks about spin echo EPIs but I wonder whether it
>>>> would also work with gradient echo functional bold EPIs? I mainly work
>>>> with non human primates and one constant problem there are apparent
>>>> motion artifacts caused by changes to the magnetic field when ever the
>>>> animal changes his position slightly. So ideally I would like to run a
>>>> dual echo EPI sequence so I can adjust each functional 3D volume for
>>>> susceptibility changes. (I already use prelude and fugue but that only
>>>> accounts fort the "static" deformations and does not account for the
>>>> inter run changes in the susceptibility profile due to the animal
>>>> changing the position of its body inside the magnet's bore).
>>>> .
>>>> best regards
>>>> sebastian
>>>>
>>>> On Sep 26, 2012, at 07:04 , Jesper Andersson wrote:
>>>>
>>>>> Hi there,
>>>>>
>>>>> topup is now openly supported. If you go to what's new in FSL 5 you'll
>>>>> find links to the documentation (maybe that's what you did?).
>>>>>
>>>>> Jesper
>>>>>
>>>>> On 26 Sep 2012, at 16:01, Marenco, Stefano (NIH/NIMH) [E] wrote:
>>>>>
>>>>>> Is topup now part of the official distribution? I could not find it
>>>>>> among the main list of tools, although when I search for topup I can
>>>>>> find a Users guide and an example. Is this now openly supported or
>>>>>> still in testing?
>>>>>>
>>>>>> Stefano Marenco, MD
>>>>>> NIMH/CBDB
>>>>>> 10 Center Drive, Bldg 10 room 3C103
>>>>>> Bethesda MD 20892
>>>>>> Tel 301 435-8964
>>>>>> Fax 301 480-7795
>>>>>> Email: [log in to unmask]
>>>>>>
>>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Jesper Andersson [mailto:[log in to unmask]]
>>>>>> Sent: Wednesday, September 26, 2012 3:58 AM
>>>>>> To: [log in to unmask]
>>>>>> Subject: Re: [FSL] TopUp --datain
>>>>>>
>>>>>> Hi Michel,
>>>>>>
>>>>>> as a little addendum to Saads reply I should stress that topup can
>>>>>> only be used if the data was acquired in a way that allows it. I.e. if
>>>>>> data was for example acquired twice with opposing phase-encode blips.
>>>>>> Data that was acquired without planning to use topup is unlikely to be
>>>>>> possible to process with topup.
>>>>>>
>>>>>> This is equivalent to other types of distortion correction where if
>>>>>> for example one has not acquired fieldmaps one cannot use fieldmap
>>>>>> based corrections.
>>>>>>
>>>>>> Good luck Jesper
>>>>>>
>>>>>>
>>>>>> On 26 Sep 2012, at 09:49, Michel Thiebaut de Schotten wrote:
>>>>>>
>>>>>>> Dear experts,
>>>>>>>
>>>>>>> This morning I tried to run topup for the first time, to correct
>>>>>>> distortion on a DWI dataset.
>>>>>>> Following the help menu I wrote the following command
>>>>>>>
>>>>>>> topup --imain=DWI.nii --datain=bvecs
>>>>>>>
>>>>>>> But the following error occurred:
>>>>>>> TopupFileIO:: msg=TopupDatafileReader:: error reading file bvecs
>>>>>>> What is supposed to be the --datain if not the Bvec?
>>>>>>>
>>>>>>> Thanks for your help an the new release of FSL,
>>>>>>> Kind regards,
>>>>>>>
>>>>>>> Michel
>>>>>>>
>>>>>>
>>>>
>>>> --
>>>> Sebastian Moeller
>>>>
>>>> telephone: +1-626-325-8598 /+1-626-395-6523 / +1-626-395-6616
>>>> fax: 626-395-8826
>>>> German GSM: +49 - 15 77 - 1 90 31 41
>>>> mobile: +1-626-325-8598
>>>> +1-626-807-5242
>>>> US CDMA: +1-626-807-5242
>>>> [log in to unmask]
>>>>
>>>> Division of Biology
>>>> MC 114-96
>>>> California Institute of Technology
>>>> 1200 East California Boulevard
>>>> CA 91125, Pasadena
>>>> USA
>>
>> --
>> Sebastian Moeller
>>
>> telephone: +1-626-325-8598 /+1-626-395-6523 / +1-626-395-6616
>> fax: 626-395-8826
>> German GSM: +49 - 15 77 - 1 90 31 41
>> mobile: +1-626-325-8598
>> +1-626-807-5242
>> US CDMA: +1-626-807-5242
>> [log in to unmask]
>>
>> Division of Biology
>> MC 114-96
>> California Institute of Technology
>> 1200 East California Boulevard
>> CA 91125, Pasadena
>> USA
>>
--
Sebastian Moeller
telephone: +1-626-325-8598 /+1-626-395-6523 / +1-626-395-6616
fax: 626-395-8826
German GSM: +49 - 15 77 - 1 90 31 41
mobile: +1-626-325-8598
+1-626-807-5242
US CDMA: +1-626-807-5242
[log in to unmask]
Division of Biology
MC 114-96
California Institute of Technology
1200 East California Boulevard
CA 91125, Pasadena
USA
|