Bob,
I also use the blip -1 direction with a Siemens 3T Trio, but will
mention that it is easier to verify the correct blip direction by
scanning a water phantom with an air bubble, so there is a distinct
change in field at the air/water interface, and plotting the signal
through the air bubble in fieldmap-corrected and uncorrected images.
Below is a scrap of code I use to do the plotting, averaging 10 images
together to improve SNR.
Jim
%
% JNL. Script to pick up SPM volume and graph results at bubble edge
%
% Jan 9, 2013, modify this script to see how unwarping changes things
% run this on fmri-raid, sitting in fieldmap3 directory, use spm5_source
% and m9 matlab
%
addpath('/export/fmri2/backup/spm5_source');
slice=17;
uncoravg = zeros(35,64,64);
coravg = zeros(35,64,64);
names = {'01', '02', '03', '04', '05', '06', '07', '08', '09', '10'};
for i=1:length(names);
uncorname = strcat('run3/af13-0004-000',names{i},'-0000',names{i},'-01.img');
% P = char('run3/uaf13-0004-00001-000001-01.img');
P = uncorname;
V = spm_vol(P);
for k=1:V.dim(3),
d = V.dim(1:2);
M = spm_matrix([0 0 k]);
img = spm_slice_vol(V,M,d,0);
uncorimg(k,:,:)=img;
end
uncoravg=uncoravg+uncorimg;
corname = strcat('run3/uaf13-0004-000',names{i},'-0000',names{i},'-01.img');
P =corname;
V = spm_vol(P);
for k=1:V.dim(3),
d = V.dim(1:2);
M = spm_matrix([0 0 k]);
img = spm_slice_vol(V,M,d,0);
corimg(k,:,:)=img;
end
coravg = coravg + corimg;
end
figure; imagesc(squeeze(uncoravg(17,:,:))); colormap(gray);
figure; imagesc(squeeze(coravg(17,:,:))); colormap(gray);
figure; plot(squeeze(uncoravg(17,39,:))); hold on;
plot(squeeze(coravg(17,39,:)),'r');
On 1/8/13, Bob Spunt <[log in to unmask]> wrote:
> ... apologies for the re-post, still looking for advice ...
>
> Dear SPM experts,
>
> I am using the SPM fieldmap toolbox to compute VDMs/ unwarp EPIs (oblique
> slice prescription, about -25 degrees) based on magnitude images and a
> presubtracted phase image (collected using a Siemens Trio 3T). I have
> determined all protocol-specific parameters except for the blip direction.
> A colleague recommended I run it both ways (Blip +1 and Blip -1) and
> compare the resulting unwarped EPIs to determine which is the correct
> value.
>
> I have done this, however, the differences are not as clear as I had
> expected to see. Attached is the most notable difference among the two,
> showing that Blip +1 seems to smear pixels posteriorly and thus fills in
> areas of dropout (I note again that these images are acquired with a pretty
> steep slice angle). Blip -1 does the opposite (not surprisingly); moreover,
> the co-registration of the Blip -1 unwarped EPIs with the hi-res anatomical
> (not pictured) is slightly (but noticeably) better than with the Blip +1.
>
> Any advice on which is correct and/or tips on how to better evaluate the
> results? I of course want to make sure I'm using these maps appropriately -
> thank you in advance for any help!
>
> Cheers,
> Bob
>
> [image: Inline image 1]
>
>
>
> On Fri, Dec 21, 2012 at 12:21 PM, Bob Spunt <[log in to unmask]> wrote:
>
>> Dear SPM experts,
>>
>> I am using the SPM fieldmap toolbox to compute VDMs/ unwarp EPIs (oblique
>> slice prescription, about -25 degrees) based on magnitude images and a
>> presubtracted phase image (collected using a Siemens Trio 3T). I have
>> determined all protocol-specific parameters except for the blip
>> direction.
>> A colleague recommended I run it both ways (Blip +1 and Blip -1) and
>> compare the resulting unwarped EPIs to determine which is the correct
>> value.
>>
>> I have done this, however, the differences are not as clear as I had
>> expected to see. Attached is the most notable difference among the two,
>> showing that Blip +1 seems to smear pixels posteriorly and thus fills in
>> areas of dropout (I note again that these images are acquired with a
>> pretty
>> steep slice angle). Blip -1 does the opposite (not surprisingly);
>> moreover,
>> the co-registration of the Blip -1 unwarped EPIs with the hi-res
>> anatomical
>> (not pictured) is slightly (but noticeably) better than with the Blip +1.
>>
>> Any advice on which is correct and/or tips on how to better evaluate the
>> results? I of course want to make sure I'm using these maps appropriately
>> -
>> thank you in advance for any help!
>>
>> Cheers,
>> Bob
>>
>> [image: Inline image 1]
>>
>
|