Hi Katie

The issue I see pertains to segmentation, since the mixture model expects the skull. If you export the batch as script, set to 0 the expected number of Gaussians, this should work (from the GUI it will change your 0, back to a value). Then save tissue classes as usual to be imported for DARTEL. 

Cyril

----- Reply message -----
From: "kathryn jankowski" <[log in to unmask]>
To: "[log in to unmask]" <[log in to unmask]>
Subject: [SPM] Field map correction in SPM12
Date: Sat, Jan 28, 2017 01:26

Thank you so much! My entire lab greatly appreciates your very kind
help. Thank you!

May I ask a follow-up question? A colleague had betted (skull-stripped)
her EPI images during preprocessing. Our lab was wondering if it is
problematic to bet EPI images if you use DARTEL. Some colleagues had
heard that betting improves coregistration, while others were concerned
about betting if DARTEL expects skull. In other words, is betting
problematic since DARTEL uses skull as one of the tissue types for
segmentation?

Thank you!
Katie

---
Kathryn F Jankowski, MS
Developmental psychology doctoral student
Autism Speaks Weatherstone Predoctoral Fellow
University of Oregon Autism Interest Group President
Developmental Social Neuroscience Lab, University of Oregon


On 2017/01/26 05:38, Guillaume Flandin wrote:
> Dear Katie,
>
> On 25/01/17 19:38, Katie Jankowski wrote:
>> Hello, We would like to do field map correction in SPM12 for our EPI
>> data (double echo field map acquisition with a Siemens Skyra), but
>> there is some confusion/debate about some of the methods. I have
>> searched online and on the SPM12 manual but I still cannot find the
>> answers to the following questions:
>>
>> 1) What is the distinction between the "Precalculated Phase and
>> Magnitude Maps" and "Calculate VDM" modules in SPM12? Some people
>> have told me to use one, others to use another.
>
> This should be the same. Older versions of SPM had four separate
> modules
> for FieldMap (presubstracted, re/im, phases, precomputed) with a lot of
> duplicated code but this has been changed such that there is now a
> single "Calculate VBM" with four options inside (entry "field map").
>
>> 2) Do you need to bet (skull strip) the magnitude field map? (Do you
>> need to, or should you, skull strip the EPI and anatomical images, if
>> you are using DARTEL segmentation and normalization?)
>
> I don't think so as the FieldMap toolbox will itself perform
> segmentation to calculate a brain mask.
>
>> 3) Finally, when I was running a SPM12 script which included the
>> following modules: "Calculate VDM", "Realign and Unwarp", and
>> "Coregister", a Fieldmap Toolbox gui popped up while running the
>> script. This toolbox, similar to what can be seen on
>> http://www.mccauslandcenter.sc.edu/crnl/tools/fieldmap, included
>> values that did not match the short TE, long TE, and EPI Readout Time
>> that I included in my modified pm_defaults file (which I inputted
>> into the "Calculate VDM" module). Does anyone know what is calling
>> this pop-up, how to prevent it from being called/how to change the
>> values in the pop-up to match the values I included in the
>> pm_defaults file, or a way to confirm that SPM is referencing the
>> values in the pm_defaults file and not in this pop-up?
>
> This interface shouldn't appear and should actually be entirely removed
> from the SPM source code. I'll look into this.
>
> Best regards,
> Guillaume.
>
> --
> Guillaume Flandin, PhD
> Wellcome Trust Centre for Neuroimaging
> University College London
> 12 Queen Square
> London WC1N 3BG