There are no intermediate files outputs in the pipeline unless specified to do so. Therefore, specified outputs are highlighted in BOLD and RED
GLOBAL VARIABLES:
sulcal_xfm-
list of the xfms that get the images from reslice to sulcal space
output-
list of names that you want the outputs to be named
reslice_xfm-
list of xfms that get the images from 305 to reslice space
air_input-
list of air files that get the images from reslice to 305 space
mask_input-
list of mask files that will be used to get hemispheres from the whole
brain segmented images
timelocked_xfm- list of xfms that get
the images from sulcal space to timelocked sulcal space
seg_input-
list of whole brain segmented images
path-
the path to where all the above .list files live
MASK VOLUME:
This module takes 2 inputs: The hemisphere mask (R or L) and the segmented
image in reslice space.
Location of files:
Hemisphere
masks: /loni/edevel/leonard/{subjectnumber}/HEMISPHERES/{subjectnumber}_reslice_mask_1mm_addgreycsf_{L
or R}_nocereb.mnc.gz
Segmented image: /loni/edevel/leonard/{subjectnumber}/{subjectnumber}_reslice_seg_rf_200it_brain_addgrey.mnc.gz
Function:
Multiplies the segmented image by either mask.
Outputs:
R and L hemisphere of segmented image.
MINC TO ANALYZE IMAGE:
This module takes 1 input: The output .mnc from the MASK
VOLUME module.
Note:
Uses the command 'mnc2ana' to change the mnc file to a img file instead
of the 'rawtominc' command that is used in the scripts.
Function:
Changes an image from .mnc format to .img
format.
Outputs:
Image in .img format
RESLICE AIR:
This module takes 3 inputs: An air file, a value for interpolation, and
the output .img from the MINC TO ANALYZE
IMAGE module.
Location
of files: Air files:
/loni/edevel/leonard/{subjectnumber}/{subjectnumber}_avg_305_AIR1ord.air
Values:
Interpolation model: value is set to "0" which codes for 'nearest neighbour'
Function:
Uses the AIR program to put the image through
a specified transformation according to the input .air file. In this case,
it is put into 305 space.
Outputs:
Segmented image in 305 space in .img format
ANALYZE IMAGE TO MINC: This
module takes 2 inputs: A model volume and the output from the RESLICE
AIR module.
Location
of files: Model volume:
/loni/edevel/LEONARD_THICKNESS/LISTS/model_305.mnc
Function:
Changes an image from .img format to .mnc format. Uses the model volume
to set the dimension size, dim order, and step of the output .mnc
Outputs:
Segmented image in 305 space in .mnc format
MINC RESAMPLE: 305 to Reslice.
This module takes 3 inputs: The transform xfm, a model volume, and the
output from the ANALYZE IMAGE TO MINC module.
Location
of files: Transform xfm:
/loni/edevel/LEONARD_HEMISPHERE_SULCAL/CONCAT_XFMS/{subjectnumber}_original_305_to_reslice.xfm
Model volume: /icbm/MNI_DATA/templates/icbm_template_1.00mm.mnc
Note:
The check boxes for the short flag and nearest neighbour flag are turned
on in this module.
Function:
Changes the image from 305 to reslice space. Uses the model volume to set
dimension size, dim order, and step of the output .mnc
Outputs:
Segmented image in reslice space in .mnc format
MINC RESAMPLE: Reslice to Sulcal.
This module takes 3 inputs: The transform xfm, a model volume, and the
output from the MINC RESAMPLE: 305 to Reslice
module.
Location
of files: Transform xfm:
/loni/edevel/LEONARD_HEMISPHERE_SULCAL/FINAL_TAGS/{subjectnumber}_avg_YALE176_sulcal.xfm
Model volume: /icbm/MNI_DATA/templates/icbm_template_1.00mm.mnc
Note:
The check boxes for the short flag and nearest neighbour flag are turned
on in this module.
Function:
Changes the image from reslice space to sulcal space. Uses the model
volume to set dimension size, dim order, and step of the output .mnc
Outputs:
Segmented image in sulcal space in .mnc format
MINC RESAMPLE: Sulcal to timelocked_sulcal.
This module takes 3 inputs: The transform xfm, a model volume, and the
output from the MINC RESAMPLE: Reslice
to Sulcal module.
Location
of files: Transform xfm:
/loni/edevel/LEONARD_THICKNESS/REAL_AND_FAKE_TIMELOCKED_XFMS/timelock_{subjectnumber}_to_*.xfm
Model volume: /icbm/MNI_DATA/templates/icbm_template_1.00mm.mnc
Note:
The check boxes for the short flag and nearest
neighbour flag are turned on in this module. Some of the xfms used are
fake xfms because images scanned at
Time 1 do not need to be timelocked. Time 1 scans go through the fake xfm
transformation which in reality, does not transform the file at all.
Uses the model volume to set dimension size, dim order, and step of the
output .mnc
Function:
Changes the image from sulcal space to timelocked, sulcal space.
Outputs:
Segmented image in timelocked sulcal space in .mnc format. Outputs to /loni/edevel/LEONARD_THICKNESS/SULCAL
GLOBAL VARIABLES:
output_thickness_ucf- will name
the output thickness ucf the same name as the subject files listed in the
segmented_subject_minc list + _thickness.ucf
subject_ucf-
a list of the blend ucfs for each subject's hemisphere
output_thickness_minc- will name
the output thickness mnc the same name as the subject files listed in the
segmented_subject_minc list + _thickness.mnc
segmented_subject_minc- a list of the timelocked
sulcal segmented hemisphere images that will be the input of this pipeline
MINC MATH: White matter clamped.
This module takes 1 input: the segmented image that is in timelocked, sulcal
space.
Location
of files: Segmented image
in timelocked, sulcal space:
/loni/edevel/LEONARD_THICKNESS/SULCAL/{subjectnumber}_timelocked_sulcal_seg_noclamp_(L
or R).mnc
Note:
The range 256, 258 was used to clamp white matter.
Function:
The range specifies what signal value (tissue) will be clamped to 1. In
this case white matter is clamped at 1 while all other signal values in
mnc are 0.
Outputs:
Segmented image with white matter clamped at 1, all other tissues/background
are 0.
MINC MATH: White matter x2.
This module takes 1 input: the output from the MINC
MATH: White matter clamped module.
Note:
The module uses one constant value that is
set at '2'. The multiply checkbox is checked on.
Function:
Multiplies the white matter clamped segmented image by the constant '2'
which makes all white matter have a value of '2' and the rest '0'.
Outputs:
Segmented image with white matter clamped at 2, all other tissues/background
are 0.
MINC MATH: Gray matter.
This module takes 1 input: the segmented image that is in timelocked, sulcal
space.
Location
of files: Segmented image
in timelocked, sulcal space:
/loni/edevel/LEONARD_THICKNESS/SULCAL/{subjectnumber}_timelocked_sulcal_seg_noclamp_(L
or R).mnc
Note:
The range 500, 520 was used to clamp gray matter.
Function:
The range specifies what signal value (tissue) will be clamped to 1. In
this case gray matter is clamped at 1 while all other signal values in
mnc are 0.
Outputs:
Segmented image with white matter clamped at 1, all other tissues/background
are 0.
MINC MATH: Add gray and white.
This module takes 2 inputs: the output from MINC
MATH: White matter x2 and MINC
MATH: Gray matter
Function:
Adds the mnc with white matter clamped at 2 and mnc with gray matter clamped
at 1 so we have a mnc with gray=1, white=2, rest =0.
Outputs:
Segmented image with gray matter clamped at 1, white matter clamped at
2, and the rest (csf and background) clamped at 0.
AUTOCROP VOLUME1: This
module takes 1 input: the output from the MINC
MATH: Add gray and white module.
Function:
Crops out extra parts of background around the brain in the segmented,
clamped image. This makes processing go faster.
Outputs:
Cropped segmented image with gray matter clamped at 1, white matter clamped
at 2, and the rest (csf and background) clamped at 0.
FLIP VOLUME (for right hemisphere ONLY):
This module takes 1 input: the output from the AUTOCROP
VOLUME1 module.
Note:
This is done to right hemisphere only. Floatseg module fails if the hemisphere
is in right hem space so it the right hem is flipped to left hem space.
The right hem also needs to be flipped into left hem space for analysis.
Function:
Takes the input image and flips it into opposite space.
Outputs:
An duplicate image of the original input that is in the exact opposite
space of that input.
MINC MATH: clamp background.
This module takes 1 input: the output from the AUTOCROP
VOLUME1 module.
Note:
The constant value is set to '0.5' and the
'Less than' checkbox is checked on.
Function:
Everything with a signal value of less than 0.5 will be clamped to 1. In
this case, csf and background will be clamped to 0.
Outputs:
Segmented image with csf and background clamped at 1 and the rest (white
and gray matter) at 0.
MINC MATH: mult by 10.
This module takes 1 input: the output from the MINC
MATH: clamp background module.
Note:
The constant value is set to '10' and the 'multiply' checkbox is checked
on.
Function:
Multiplies the input by 10 which results in all csf and background having
the value of 10 and the rest (white and gray) at 0.
Outputs:
Segmented image with csf and background clamped at 10 and the rest (white
and gray matter) at 0.
MINC MATH: add.
This module takes 2 inputs: the output from MINC
MATH: mult by 10 and AUTOCROP
VOLUME1.
Function:
Adds the image with gray and white clamped at 1 and 2, respectively, and
image with csf and background clamped at 10 together.
Outputs:
Segmented image with csf and background at 10, gray at 1, and white at
2.
MINC RESAMPLE: Resample back to to be like
previous. This module takes 2 inputs:
the output from MINC MATH: add
and MINC MATH: Add gray and white.
Function:
Resamples the cropped segmented image (background=10, gray=1, white=2)
to it's original dimensions.
Outputs:
Segmented image with original dimensions (background=10, gray=1, white=2,
padding=0)
MINC RESAMPLE: Resample to have floats and
thirds. This module takes 1 input: the
output from MINC RESAMPLE: Resample back
to to be like previous.
Note:
The 'float' and 'nearest neighbour' checkboxes are turned on. The Xdim,
Ydim, Zdim values are 543, 651, 543, respectively. The xstep, ystep, zstep
are all set to the value '0.333'.
Function:
Takes the input and forces it to have floating numbers for values, and
resamples it to have 0.333 step sizes, and user specified dimensions.
Outputs:
Segmented image with "third-size' dimensions, forced float number values,
and step sizes of 0.333 (one-third of millimeter).
AUTOCROP VOLUME2:
This module takes 1 input: the output from the MINC
RESAMPLE: Resample to have floats and thirds
module.
Function:
Takes the input and crops the excess background out of the image. This
makes FLOATSEG work
faster.
Outputs:
Cropped image
FLOATSEG: This
module has 1 input: the output from the AUTOCROP
VOLUME2 module.
Note:
The value for 'Max thickness' is set at 10. The white matter 'min' and
'max' values are set at 1.5 and 2.5 respectively. This function takes a
LONG
time.
Function:
Takes the input and outputs a thickness image.
Outputs:
Thickness mnc to /loni/edevel/LEONARD_THICKNESS/SULCAL
MINC MATH: clamp gray.
This module takes 1 input: the output from the AUTOCROP
VOLUME2 module.
Note:
The constant values are set at 0.98 and 1.1.
Function:
Clamps the gray matter signal values making their value equal to 1 and
the rest 0.
Outputs:
Segmented image file with gray matter = 1
and the rest = 0.
MASK VOLUME: gray only in thickness.mnc.
This module takes 2 inputs: the outputs from FLOATSEG
and MINC MATH: clamp gray.
Function:
Multiplies the thickness mnc with the gray matter mask to obtain a gray
matter only thickness map.
Outputs:
Thickness mnc that is gray matter only.
SOFTMEAN onto UCF DBL: double thickness.mnc
and make ucfs. This module takes 1 input:
The putput from MASK VOLUME: gray only
in thickness.mnc.
Note:
The kernel value is set to 15. The min and max thickness values are set
to 0.3 and 10 respectively.
Function:
Takes the input and runs a smoothing kernel over the surface averaging
cortical thickness at points on the hemisphere. Outputs each point's average
to the 4th dimension of a ucf file.
Outputs:
Ucf files for each subject's hemisphere with average cortical thickness
in the 4th dimension. Outputs to /loni/edevel/LEONARD_THICKNESS/SULCAL