Hi Matthew, 

Yes, the mask is generated from the T1 structural. 

Your steps are extremely helpful and should solve my problem! 

Many thanks of your help.

BW

Clare
From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of Matthew Webster [[log in to unmask]]
Sent: 23 November 2015 16:30
To: [log in to unmask]
Subject: Re: [FSL] [FSL] Convert data type to FLOAT32

Hello Clare,
                      Was the mask generated from the T1 structural image, or was it generated from a different source image? Passing a pre-threshold mask to FEAT, requires that the mask is in the same space as the functional data - i.e. it must be manually registered before FEAT is run. It is unlikely the mask can be directly registered to the functional data. Instead something like the following process must be used:

1)
It is important that the segmentation image has the same geometry as the structural it was generated from. Cropping and resampling will prevent the rest of the steps from working correctly
2)
Register the functional data to the structural image used to generate the segmentation - save the transformation matrix
3)
Invert the transformation ( e.g. using FSL's convert_xfm tool ) to obtain the reverse transformation from structural to functional space
4)
Apply this reverse transformation to the mask to obtain the mask in functional space - you may want to re-binarise this mask to remove any interpolation effects.
5)
This mask is now suitable for use with FEAT.

Hope this helps,
Kind Regards
Matthew
Hi Paul,

Many thanks for your reply. I was attempting to register functional to structural T1 and include the mask in a single FEAT analysis. This seemed to work fine as the warning occurred only at post stats. Perhaps it needs to be completed in a two-step process: do the registration first then the ROI analysis separately? I'll give that a go unless I've misunderstood.

Thanks again,

Clare 

Sent from my iPhone

On 23 Nov 2015, at 15:55, paul mccarthy <[log in to unmask]> wrote:

Hi Clare,

Your mask images need to be defined in functional space. Have you registered your functional images to their structural counterparts? If so, you will be able to use the stored transformations to move your existing structural mask images into functional space.

Cheers,

Paul

On 23 November 2015 at 15:34, Clare Loane <[log in to unmask]> wrote:
Hello again,

I wonder if the info I provided below is enough to understand why I am receiving the warning when trying to run FEAT with my mask?

Thank you in advance.

Clare
________________________________________
From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of Clare Loane [[log in to unmask]]
Sent: 20 November 2015 12:06
To: [log in to unmask]
Subject: Re: [FSL] Convert data type to FLOAT32

Dear Matthew,

Thanks for the reply. I have ran fslhd on the functional data and mask and they do not match:

filename       con_012_func1.nii.gz

sizeof_hdr     348
data_type      INT16
dim0           4
dim1           64
dim2           64
dim3           44
dim4           261
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        2.410000
pixdim5        1.000000
pixdim6        1.000000
pixdim7        58545.914062
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      0
slice_name     Unknown
slice_code     0
slice_start    0
slice_end      0
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
qto_xyz:1      -3.000000  0.000000  -0.000000  96.000000
qto_xyz:2      0.000000  2.598076  -1.500000  -44.900520
qto_xyz:3      0.000000  1.500000  2.598076  -106.426422
qto_xyz:4      0.000000  0.000000  0.000000  1.000000
qform_xorient  Right-to-Left
qform_yorient  Posterior-to-Anterior
qform_zorient  Inferior-to-Superior
sform_name     Scanner Anat
sform_code     1
sto_xyz:1      -3.000000  0.000000  0.000000  96.000000
sto_xyz:2      0.000000  2.598076  -1.500000  -44.900520
sto_xyz:3      0.000000  1.500000  2.598076  -106.426422
sto_xyz:4      0.000000  0.000000  0.000000  1.000000
sform_xorient  Right-to-Left
sform_yorient  Posterior-to-Anterior
sform_zorient  Inferior-to-Superior
file_type      NIFTI-1+
file_code      1
descrip
aux_file




filename       con_012_seg.nii.gz

sizeof_hdr     348
data_type      UINT16
dim0           3
dim1           185
dim2           207
dim3           209
dim4           1
dim5           1
dim6           1
dim7           1
vox_units      mm
time_units     Unknown
datatype       512
nbyper         2
bitpix         16
pixdim0        0.000000
pixdim1        1.000000
pixdim2        1.000000
pixdim3        1.000000
pixdim4        0.000000
pixdim5        0.000000
pixdim6        0.000000
pixdim7        0.000000
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      0
slice_name     Unknown
slice_code     0
slice_start    0
slice_end      0
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     Aligned Anat
qform_code     2
qto_xyz:1      -1.000000  -0.000000  -0.000000  92.965897
qto_xyz:2      0.000000  1.000000  -0.000000  -79.256493
qto_xyz:3      -0.000000  0.000000  1.000000  -103.286942
qto_xyz:4      0.000000  0.000000  0.000000  1.000000
qform_xorient  Right-to-Left
qform_yorient  Posterior-to-Anterior
qform_zorient  Inferior-to-Superior
sform_name     Scanner Anat
sform_code     1
sto_xyz:1      -1.000000  -0.000000  0.000000  92.965897
sto_xyz:2      -0.000000  1.000000  -0.000000  -79.256493
sto_xyz:3      0.000000  0.000000  1.000000  -103.286942
sto_xyz:4      0.000000  0.000000  0.000000  1.000000
sform_xorient  Right-to-Left
sform_yorient  Posterior-to-Anterior
sform_zorient  Inferior-to-Superior
file_type      NIFTI-1+
file_code      1
descrip
aux_file

I also want to use the individual T1 and the fslhd info for this is:

filename       con_012_mprage_brain.nii.gz

sizeof_hdr     348
data_type      FLOAT32
dim0           3
dim1           174
dim2           192
dim3           192
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
cal_max        0.0000
cal_min        0.0000
scl_slope      1.000000
scl_inter      0.000000
phase_dim      0
freq_dim       0
slice_dim      0
slice_name     Unknown
slice_code     0
slice_start    0
slice_end      0
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
qto_xyz:1      -1.000000  0.000000  -0.000000  87.713997
qto_xyz:2      0.000000  0.982935  -0.183951  -65.430283
qto_xyz:3      0.000000  0.183951  0.982935  -122.346207
qto_xyz:4      0.000000  0.000000  0.000000  1.000000
qform_xorient  Right-to-Left
qform_yorient  Posterior-to-Anterior
qform_zorient  Inferior-to-Superior
sform_name     Scanner Anat
sform_code     1
sto_xyz:1      -1.000000  0.000000  0.000000  87.713997
sto_xyz:2      0.000000  0.982935  -0.183951  -65.430283
sto_xyz:3      0.000000  0.183951  0.982935  -122.346207
sto_xyz:4      0.000000  0.000000  0.000000  1.000000
sform_xorient  Right-to-Left
sform_yorient  Posterior-to-Anterior
sform_zorient  Inferior-to-Superior
file_type      NIFTI-1+
file_code      1
descrip        FSL5.0
aux_file

I hope this is helpful? Many thanks again,

BW

Clare

________________________________________
From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of Matthew Webster [[log in to unmask]]
Sent: 19 November 2015 13:44
To: [log in to unmask]
Subject: Re: [FSL] Convert data type to FLOAT32

Hello,
           This probably isn't an issue with data-type - the command "/usr/local/fsl/bin/fslmaths thresh_zstat1 -mas /Users/.../Desktop/SCANS/Masks/seg_mask.nii.gz thresh_zstat1" is failing because the mask has a different geometry to the functional data. From the output below, it looks like your mask is 185x207x209 in 1mm resolution, does this match the functional data? Can you let me know the results of running fslhd on both the functional data and the mask?

Kind Regards
Matthew

> Dear All,
>
> I want to conduct fMRI FEAT analyses at the individual level with subject specific MRIs and manual segmentation masks. The masks were created in ITK-snap and INT16 type, not FLOAT32 as the other images are.
>
> data_type      INT16
>
> dim1           185
>
> dim2           207
>
> dim3           209
>
> dim4           1
>
> datatype       4
>
> pixdim1        1.000000
>
> pixdim2        1.000000
>
> pixdim3        1.000000
>
> pixdim4        0.000000
>
> cal_max        0.0000
>
> cal_min        0.0000
>
> file_type      NIFTI-1+
>
>
> Therefore, when I am running the feat, I get an error at post stats:
>
> WARNING:: Inconsistent orientations for individual images in pipeline!
>           Will use voxel-based orientation which is probably incorrect - *PLEASE CHECK*!
>
> Image Exception : #3 :: Attempted to multiply images/ROIs of different sizes
> libc++abi.dylib: terminating with uncaught exception of type RBD_COMMON::BaseException
>
> FATAL ERROR ENCOUNTERED:
> COMMAND:
> /usr/local/fsl/bin/fslmaths thresh_zstat1 -mas /Users/.../Desktop/SCANS/Masks/seg_mask.nii.gz thresh_zstat1
> ERROR MESSAGE:
> child killed: SIGABRT
> END OF ERROR MESSAGE
> child killed: SIGABRT
>     while executing
> "if { [ catch {
>
> for { set argindex 1 } { $argindex < $argc } { incr argindex 1 } {
>     switch -- [ lindex $argv $argindex ] {
>
>       -I {
>           incr arginde..."
>     (file "/usr/local/fsl/bin/feat" line 310)
> Error encountered while running in main feat script, halting.
> child killed: SIGABRT
>
>
> I am as sure that the images are labelled correctly, so assume the data type is the problem.
>
> Does this sound correct? And if so, how do I safely convert the mask to the correct data type?
>
> Many thanks in advance.
>
> Clare