Print

Print
Hi Vasudev,

FEAT scales the functional data as part of its preparation for model fitting. The magnitude of the fMRI signal is arbitrary - what is important is how the signal changes over time.

Paul
On Feb 21 2020, at 1:00 pm, Dev vasu <[log in to unmask]> wrote:
Dear Sir,

Thanks for your reply, Now i found out the issue, my mask is actually made out of
ICBM 2009c Nonlinear Symmetric - 1×1x1mm template

Now i have transformed my mask into MNI152 2mm template,  I have a question, i extracted time series using

fslmeants -i filtered_func_data.nii.gz -o Visual.txt -m Visual_thr_func.nii.gz

the Visual_thr_func.nii.gz is generated after transforming Visual_thr_mask.nii.gz into functional space.

The time series file that i am getting ( see attachment) is huge, can such high values for time series are possible ?.

Thanks
Vasudev


On Fri, Feb 21, 2020 at 1:31 PM paul mccarthy <[log in to unmask]> wrote:
Hi Vasudev,

You may have missed my most recent email:

Actually, I was slightly wrong - your mask image is aligned to the MNI152 template. However, it does not have the same resoluion as the MNI152 template you used in registration (2mm I presume).

Your mask needs to have the same dimensions and orientation as the standard space image you used for registration.

Paul

On Feb 21 2020, at 12:24 pm, Dev vasu <[log in to unmask]> wrote:
Dear Sir,

I am annexing Visual_thr_mask.nii.gz file,it is actually in MNI152 space.

Thanks
Vasudev

On Fri, Feb 21, 2020 at 1:01 PM paul mccarthy <[log in to unmask]> wrote:
Hi Vasudev,

I've just noticed that the file you attached to the beginning of this email thread (Visual_thr_mask.nii.gz ) is not in MNI space - I'm guessing it is in the space of your structural image. This command: 

applywarp -i Visual_thr_mask.nii.gz -r reg/example_func.nii.gz -o Visual_func --postmat=reg/highres2example_func.mat -w reg/highres2standard_warp_inv

assumes that the input file is in MNI152 space.

Paul

On Feb 21 2020, at 9:41 am, Dev vasu <[log in to unmask]> wrote:
Dear Sir,

I have redone the FEAT for few subjects without nonlinear registration. (see design.fsf attached)

1. Links to registrations


- The process you used to generate highres2standard_warp_inv  is valid

This is the process that i have used to generate highres2standard_warp_inv

convert_xfm -omat highres2example_func.mat -inverse example_func2highres.mat
invwarp --ref==highres --warp=highres2standard_warp --out=highres2standard_warp_inv



and the output mask still has problem.








On Thu, Feb 20, 2020 at 11:00 AM paul mccarthy <[log in to unmask]> wrote:
Hi Vasudev,

No - your functional and structural data should be in the resolution they were acquired in respectively, and should match the resolution of the images you used in the registration process.  The dimensions you have quoted look like reasonable values for T1/functional EPI data.

To solve your problem, you need to identify the specific point of failure, and focus on fixing that. Refer to the list of steps in my previous email.

In your other email thread, it looked to me like the non-linear registration was not ideal - this could be caused by a poor brain extraction, or an atypical brain. In your most recent screenshot it looks like something more fundamental has gone wrong - I would hazard a guess that one of the files you have used is incorrect.

Paul
On Feb 20 2020, at 9:43 am, Dev vasu <[log in to unmask]> wrote:
Dear Sir,

- that all of the input files have the correct dimensions , is it necessary that dimensions of all functional and structural files should be same ?

The Dimensions of all structural files is
sizeof_hdr     348
data_type      FLOAT32
dim0           3
dim1           160
dim2           256
dim3           256
dim4           1
dim5           1
dim6           1
dim7           1
vox_units      mm
time_units     s
datatype       16
nbyper         4
bitpix         32
pixdim0        0.000000
pixdim1        1.000000
pixdim2        1.000000
pixdim3        1.000000
pixdim4        1.000000
pixdim5        0.000000
pixdim6        0.000000
pixdim7        0.000000
vox_offset     352

Dimensions of all Functional files

sizeof_hdr     348
data_type      INT16
dim0           4
dim1           64
dim2           64
dim3           46
dim4           300
dim5           1
dim6           1
dim7           1
vox_units      mm
time_units     s
datatype       4
nbyper         2
bitpix         16
pixdim0        0.000000
pixdim1        3.000000
pixdim2        3.000000
pixdim3        3.000000
pixdim4        2.500000
pixdim5        1.000000
pixdim6        0.253332
pixdim7        47455.246094
vox_offset     352
cal_max        0.0000
cal_min        0.0000
scl_slope      1.000000
scl_inter      0.000000
phase_dim      0
freq_dim       0
slice_dim      3
slice_name     alternating_increasing_2
slice_code     5
slice_start    0
slice_end      45
slice_duration 0.000000
time_offset    0.000000
intent         Unknown
intent_code    0
intent_name    
intent_p1      0.000000
intent_p2      0.000000
intent_p3      0.000000
qform_name     Scanner Anat
qform_code     1





On Thu, Feb 20, 2020 at 10:26 AM paul mccarthy <[log in to unmask]> wrote:
Hi Vasudev,

As I said in your other thread, this is due to a poor registration. Double check the following:

- that all of the input files have the correct dimensions
- The structural brain extraction is good
- The func-struc registration is good
- The struc-standard registration is good
- The process you used to generate highres2standard_warp_inv  is valid

Paul


On Feb 19 2020, at 8:09 pm, Dev vasu <[log in to unmask]> wrote:
Dear all,

I have done functional preprocessing using FEAT and i would like to perform seed based correlation for visual cortex seed regions, my seed is defined in MNI152 space.

The seed and example_func.nii.gz are completely mismatched ( see picture ) , I have used Visual_thr_mask.nii.gz as seed ( see attachment) and tried to transform by doing
applywarp -i Visual_thr_mask.nii.gz -r reg/example_func.nii.gz -o Visual_func 
--postmat=reg/highres2example_func.mat -w reg/highres2standard_warp_inv

I dont understand what is the problem with my approach-could you please provide
me some input.




To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1