NAMIC Wiki - User contributions [en]

Smoothing multi-label brain segmentations

2010-01-27T23:12:29Z

Ginger: /* Prepare smoothed label maps for separated labels */

== Input: (Label.nrrd) a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

== Prepare smoothed label maps for separated labels ==

* Tool: smoothing label map slicer3 module (SLM)
* Individual label maps:

** ICC: threshold Label.nrrd at all non-zero places, use default parameters in SLM to smooth the result, get binary label map

ICC.nrrd

** Smooth label 5,6,11-20, use SLM, take Label.nrrd as input, choose corresponding label to be smoothed, one at a time. All use default parameters. Result in label maps:

BrainStem.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd

RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd

** Smooth WM: choose label 3, take Label.nrrd as input, default parameters in SLM, get

WM_smooth.nrrd

** Smooth CSF: choose label 1, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map CSF_ori.nrrd for CSF same as in Label.nrrd. Go to editor, do a closing operation on CSF_ori.nrrd (a 4-neighbor dilation --an optional smoothing using SLM --followed by a 4-neighbor erosion) to get

CSF_smooth.nrrd

** GM: choose label 2, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map GM_ori.nrrd for GM same as in Label.nrrd. Go to editor, do a 4-neighbor dilation to get

GM_Dia.nrrd

== Re-combine smoothed label maps ==

* Use the application ''Combine-sub-structures'' to generate the resulting smoothed label maps.

./Combine-sub-structures ICC.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd RGlobusPallidus.nrrd LGlobusPallidus.nrrd
RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd CSF_smooth.nrrd WM_smooth.nrrd BrainStem.nrrd GM_Dia.nrrd outputAllLabel.nrrd

Smoothing multi-label brain segmentations

2010-01-27T23:12:17Z

Ginger: /* Prepare smoothed label maps for separated labels */

== Input: (Label.nrrd) a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

== Prepare smoothed label maps for separated labels ==

* Tool: smoothing label map slicer3 module (SLM)
* Individual label maps:
** ICC: threshold Label.nrrd at all non-zero places, use default parameters in SLM to smooth the result, get binary label map
ICC.nrrd
** Smooth label 5,6,11-20, use SLM, take Label.nrrd as input, choose corresponding label to be smoothed, one at a time. All use default parameters. Result in label maps:

BrainStem.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd

RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd

** Smooth WM: choose label 3, take Label.nrrd as input, default parameters in SLM, get

WM_smooth.nrrd

** Smooth CSF: choose label 1, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map CSF_ori.nrrd for CSF same as in Label.nrrd. Go to editor, do a closing operation on CSF_ori.nrrd (a 4-neighbor dilation --an optional smoothing using SLM --followed by a 4-neighbor erosion) to get

CSF_smooth.nrrd

** GM: choose label 2, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map GM_ori.nrrd for GM same as in Label.nrrd. Go to editor, do a 4-neighbor dilation to get

GM_Dia.nrrd

== Re-combine smoothed label maps ==

* Use the application ''Combine-sub-structures'' to generate the resulting smoothed label maps.

./Combine-sub-structures ICC.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd RGlobusPallidus.nrrd LGlobusPallidus.nrrd
RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd CSF_smooth.nrrd WM_smooth.nrrd BrainStem.nrrd GM_Dia.nrrd outputAllLabel.nrrd

Smoothing multi-label brain segmentations

2010-01-27T22:58:46Z

Ginger: /* Re-combine smoothed label maps */

== Input: (Label.nrrd) a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

== Prepare smoothed label maps for separated labels ==

* Tool: smoothing label map slicer3 module (SLM)
* Individual label maps:
** ICC: threshold Label.nrrd at all non-zero places, use default parameters in SLM to smooth the result, get binary label map
ICC.nrrd
** Smooth label 5,6,11-20, use SLM, take Label.nrrd as input, choose corresponding label to be smoothed, one at a time. All use default parameters. Result in label maps:

BrainStem.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd

RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd

** Smooth WM: choose label 3, take Label.nrrd as input, default parameters in SLM, get

WM_smooth.nrrd

** Smooth CSF: choose label 1, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map CSF_ori.nrrd for CSF same as in Label.nrrd. Go to editor, do a closing operation on CSF_ori.nrrd (a 4-neighbor dilation --an optional smoothing using SLM --followed by a 4-neighbor erosion) to get

CSF_smooth.nrrd

** GM: choose label 2, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map GM_ori.nrrd for GM same as in Label.nrrd. Go to editor, do a 4-neighbor dilation to get

GM_Dia.nrrd

== Re-combine smoothed label maps ==

* Use the application ''Combine-sub-structures'' to generate the resulting smoothed label maps.

./Combine-sub-structures ICC.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd RGlobusPallidus.nrrd LGlobusPallidus.nrrd
RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd CSF_smooth.nrrd WM_smooth.nrrd BrainStem.nrrd GM_Dia.nrrd outputAllLabel.nrrd

Smoothing multi-label brain segmentations

2010-01-27T22:58:38Z

Ginger:

== Input: (Label.nrrd) a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

== Prepare smoothed label maps for separated labels ==

* Tool: smoothing label map slicer3 module (SLM)
* Individual label maps:
** ICC: threshold Label.nrrd at all non-zero places, use default parameters in SLM to smooth the result, get binary label map
ICC.nrrd
** Smooth label 5,6,11-20, use SLM, take Label.nrrd as input, choose corresponding label to be smoothed, one at a time. All use default parameters. Result in label maps:

BrainStem.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd

RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd

** Smooth WM: choose label 3, take Label.nrrd as input, default parameters in SLM, get

WM_smooth.nrrd

** Smooth CSF: choose label 1, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map CSF_ori.nrrd for CSF same as in Label.nrrd. Go to editor, do a closing operation on CSF_ori.nrrd (a 4-neighbor dilation --an optional smoothing using SLM --followed by a 4-neighbor erosion) to get

CSF_smooth.nrrd

** GM: choose label 2, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map GM_ori.nrrd for GM same as in Label.nrrd. Go to editor, do a 4-neighbor dilation to get

GM_Dia.nrrd

== Re-combine smoothed label maps ==

* Use the application ''Combine-sub-structures'' to generate the resulting smoothed label maps.

./Combine-sub-structures ICC.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd RGlobusPallidus.nrrd LGlobusPallidus.nrrd
RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd CSF_smooth.nrrd WM_smooth.nrrd BrainStem.nrrd GM_Dia.nrrd outputAllLabel.nrrd

Smoothing multi-label brain segmentations

2010-01-27T22:55:56Z

Ginger: /* Prepare smoothed label maps for separated labels */

== Input: (Label.nrrd) a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

== Prepare smoothed label maps for separated labels ==

* Tool: smoothing label map slicer3 module (SLM)
* Individual label maps:
** ICC: threshold Label.nrrd at all non-zero places, use default parameters in SLM to smooth the result, get binary label map
ICC.nrrd
** Smooth label 5,6,11-20, use SLM, take Label.nrrd as input, choose corresponding label to be smoothed, one at a time. All use default parameters. Result in label maps:

BrainStem.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd

RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd

** Smooth WM: choose label 3, take Label.nrrd as input, default parameters in SLM, get

WM_smooth.nrrd

** Smooth CSF: choose label 1, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map CSF_ori.nrrd for CSF same as in Label.nrrd. Go to editor, do a closing operation on CSF_ori.nrrd (a 4-neighbor dilation --an optional smoothing using SLM --followed by a 4-neighbor erosion) to get

CSF_smooth.nrrd

** GM: choose label 2, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map GM_ori.nrrd for GM same as in Label.nrrd. Go to editor, do a 4-neighbor dilation to get

GM_Dia.nrrd

./Combine-sub-structures ICC.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd CSF_smooth.nrrd WM_smooth.nrrd BrainStem.nrrd GM_Dia.nrrd outputAllLabel.nrrd

Smoothing multi-label brain segmentations

2010-01-27T22:54:19Z

Ginger: /* Prepare smoothed label maps for separated labels */

== Input: (Label.nrrd) a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

== Prepare smoothed label maps for separated labels ==

* Tool: smoothing label map slicer3 module (SLM)
* Individual label maps:
** ICC: threshold Label.nrrd at all non-zero places, use default parameters in SLM to smooth the result, get binary label map
ICC.nrrd
** Smooth label 5,6,11-20, use SLM, take Label.nrrd as input, choose corresponding label to be smoothed, one at a time. All use default parameters. Result in label maps:

BrainStem.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd

RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd

** Smooth WM: choose label 3, take Label.nrrd as input, default parameters in SLM, get

WM_smooth.nrrd

** Smooth CSF: choose label 1, take Label.nrrd as input, change sigma = 0 in SLM, get binary label map CSF_ori.nrrd for CSF same as in Label.nrrd. Go to editor, do a closing operation on CSF_ori.nrrd (a 4-neighbor dilation --an optional smoothing using SLM --followed by a 4-neighbor erosion) to get

CSF_smooth.nrrd

./Combine-sub-structures ICC.nrrd Cerabellum.nrrd RCaudate.nrrd LCaudate.nrrd RPutamen.nrrd LPutamen.nrrd RGlobusPallidus.nrrd LGlobusPallidus.nrrd RThalamus.nrrd LThalamus.nrrd RHippocampus.nrrd LHippocampus.nrrd CSF_smooth.nrrd WM_smooth.nrrd BrainStem.nrrd GM_smooth_Dia.nrrd outputAllLabel.nrrd

Smoothing multi-label brain segmentations

2010-01-27T22:46:10Z

Ginger: /* Prepare smoothed label maps for separated labels */

Smoothing multi-label brain segmentations

2010-01-27T22:45:42Z

Ginger: /* Prepare smoothed label maps for separated labels */

Smoothing multi-label brain segmentations

2010-01-27T22:45:33Z

Ginger:

Smoothing multi-label brain segmentations

2010-01-27T22:45:14Z

Ginger:

Smoothing multi-label brain segmentations

2010-01-27T22:38:06Z

Ginger:

== Input: a multi-label map to be smoothed ==

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:37:04Z

Ginger:

== smoothing workflow ==

Input: a multi-label map to be smoothed. With structures:

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:36:51Z

Ginger:

== smoothing workflow ==

input: a multi-label map to be smoothed. With structures:

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:36:35Z

Ginger:

== smoothing workflow ==

# input: a multi-label map to be smoothed. With structures:

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20
* Other
** Cerebellum Label 6
** BrainStem Label 5

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:35:43Z

Ginger:

== smoothing workflow ==

# input: a multi-label map to be smoothed. With structures:

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RGlobusPallidus label 17
** LGlobusPallidus label 18
** RThalamus label 11
** LThalamus label 12
** RHippocampus label 19
** LHippocampus label 20

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:33:59Z

Ginger:

== smoothing workflow ==

# input: a multi-label map to be smoothed. With structures:

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RCaudate label 17
** RCaudate label 18
** RCaudate label 11
** RCaudate label 12
** RCaudate label 19
** RCaudate label 20

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:33:44Z

Ginger:

== smoothing workflow ==

# input: a multi-label map to be smoothed. With structures:

* WM label 3
* CSF label 1
* GM(without the following specified sub-cortical structures) label 2,
* GM sub-cortical structures
** RCaudate label 13
** LCaudate label 14
** RPutamen label 15
** LPutamen label 16
** RCaudate label 17
** RCaudate label 18
** RCaudate label 11
** RCaudate label 12
** RCaudate label 19
** RCaudate label 20

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:31:50Z

Ginger:

== smoothing workflow ==

# input: a multi-label map to be smoothed. With structures:

WM label 3
CSF label 1
GM(without the following specified sub-cortical structures) label 2,
GM sub-cortical structures

# Segmentation using one of the approaches below
# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

Smoothing multi-label brain segmentations

2010-01-27T22:28:33Z

Ginger: Created page with '== smoothing workflow == # Deformable registration of the atlas to the subject to be segmented # Segmentation using one of the approaches below For State 1 scans, the workflow …'

== smoothing workflow ==

# Deformable registration of the atlas to the subject to be segmented
# Segmentation using one of the approaches below

For State 1 scans, the workflow is the following:

# Up to Affine registration of the subject to bias-corrected, skull-stripped Baseline scan
# Map the Baseline ICC mask to the State 1 space
# Skull stripping State 1 scan
# Bias-correction of State 1 scan
# Repeat up to affine registration of skull-stripped, bias-corrected State 1 scan to Baseline

== Considered segmentation approaches ==

Choice of atlas:
* Vervet atlas
* Rhesus atlas

Choice of segmentation approach:
* fluid deformable registration of a specific atlas to the subject (UNC approach)
* EM Segmenter (Kilian Pohl)
* itkEMS/ABC (Marcel Prastawa)

NA-MIC NCBC Collaboration:Measuring Alcohol and Stress Interaction

2010-01-27T22:27:54Z

Ginger: /* Documentation */

Back to [[NA-MIC_External_Collaborations|NA-MIC External Collaborations]]
==Abstract==
Alcohol abuse continues to be a major problem in the United States with an estimated 14 million adults
meeting the clinical criteria for a diagnosis of alcohol abuse or alcoholism at an estimated cost of $185
billion. Chronic alcohol exposure has been shown to impact structure and function of the brain. Since
most of what is known of alcohol's effects on the brain is based on studies of individuals who have
abused for a long time, it is unclear whether some of the effects might occur early in the abuse process.
Additionally, human studies are often complicated by many factors incuding polydrug abuse, poor
nutritional states and other medical conditions. Another complicating factor is the interaction of stress
and alcohol.

Using magnetic resonance imaging (MRI), we will examine the the effects of chronic alcohol self-
administration on brain structure and function in a group of monkeys that have been mother-reared or
nursery-reared. MRI scans will be acquired to determine whether there are differences in stucture or
functions as a result of alcohol exposure and whether rearing conditions play a role in these effects.
MRI techniques will also be used to examine how chronic alcohol consumption affects cerebral blood
flow, which is a measure of neuronal activity. Because we will conduct these measures both before
and after alcohol exposure in the same "young adult" animals, we will be able to track any changes that
might occur early in alcohol abuse which is something that cannot be accomplished in human studies.
Another goal is to adapt imaging tools and computer software that are used to measure the brains of
humans to the monkey brain. To accomplish this, we will collaborate with the National Alliance for
Medical Computing (NA-MIC), an existing NIH Center for Biomedical Computing.

We can control all variables associated with alcohol self-administration in these monkeys. They will be
monitored both before and after exposure to alcohol, thus, we will be able to monitor the progression of
any changes in structure or function that directly result from alcohol exposure. These studies will help
identify how exposure to long-term alcohol might affect how the brain works.

==Grant#==
R01AA016748

==Key Personnel==
*Wake Forest University: Jim Daunais, Bob Kraft
*Virginia Tech: Chris Wyatt, Vidya Rajagopalan, Xiaoxing(Ginger) Li, Xiaojing Long
*NA-MIC: Kilian Pohl, Sandy Wells, Andriy Fedorov

==Funding Duration==
07/15/2009-03/31/2010

==Projects==
Active projects in this collaboration are:

=== Tutorials on using EMSegmenter to segment nonhuman primate images ===
#Slicer3 based tutorial
##[http://www.na-mic.org/Wiki/index.php/Image:EMSegment_TrainingTutorial.pdf Tutorial Slide Handouts (PDF)]
##[http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Vervet Atlas]
##[http://www.na-mic.org/Wiki/images/c/c9/SlicerTutorialData_Vervet.tar.gz Example Data]
#[http://bsl.ece.vt.edu/download/NHP-EMSTut.zip older Slicer2.x based tutorial]
## [http://www.na-mic.org/Wiki/index.php/Image:EMSegment_Vervet_TrainingTutorial_Slicer2.pdf Vervet segmentation with Slicer2] (extracted from above for convenience)

=== Nonhuman Primate Atlas Development ===
Vervet Probabalistic Atlas - Adult
*[http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Vervet Probabilistic Atlas (VPA-10)]

Rhesus Probabalistic Atlas - Adult
*[[Rhesus_WMGM_Atlas|WM/GM Atlas]]
*[[Rhesus_Subcortical_Atlas|Subcortical Atlas]]

Rhesus Probabalistic Atlas - Juvenile
*[[Rhesus_Juvenile_WMGM_Atlas|WM/GM Atlas]]

Cyno Probabalistic Atlas - Adult
*[[Cyno_WMGM_Atlas|WM/GM Atlas]]

=== Segmentation ===
#[[Rhesus_ICV_Extraction | Rhesus ICV Extraction]]
#[[Rhesus_EM_Segmentation | Rhesus EM Segmentation]]

=== Documentation ===
# [[Atlas-to-subject registration with FSL and Diffeomorphic Demons, for EM segmentation.]]
# [http://wiki.na-mic.org/Wiki/index.php/Vervet_MRI_registration Atlas-to-subject registration with BRAINSFit]
# [[On the performance of Slicer 2 and Slicer3 EM segmentation modules]]
# [[Proposed revised atlas construction workflow]]
# [[Initial atlas construction workflow]] (initial workflow, obsolete)
# [[October_08_2009_Meeting:_VT-NAMIC_Revised_atlas_construction_workflow | Current vervet monkey atlas construction workflow]] (in progress)
# [[NA-MIC WFU VT Collaboration: Segmentation workflow and approaches | Considered segmentation approaches and workflow]] (in progress)
# [[Smoothing multi-label brain segmentations]]

==Publications==
==Meetings and Events Specific to this Collaboration==
* [[2007_Programming/Project_Week_MIT|2007 Summer Programming/Project Week at MIT]]
* [[AHM_2008|2008 NA-MIC AHM]]
* [[NA-MIC/Projects/External Collaboration/Measuring Alcohol and Stress Interaction|2008 NA-MIC Summer Project Week Page]]
* [[2009_Winter_Project_Week_WFU1|2009 NA-MIC Winter Project Week Page]]
* [http://wiki.na-mic.org/Wiki/index.php/Measuring_Alcohol_Stress_Interaction 2009 Summer project week Boston]
* [http://wiki.na-mic.org/Wiki/index.php/Events:VT-NAMIC-NCBC-Collaboration-meeting-2009 2009 October 7-9 VT-NAMIC collaboration meeting]

==Resource Links==
* [http://www.slicer.org 3D Slicer]
* [http://www.bsl.ece.vt.edu Bioimaging Systems Lab at VT]
* [http://pin.primate.wisc.edu/factsheets/entry/vervet Primate Info Net: Vervet factsheet]

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2010-01-05T18:45:33Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| baseline T1 data source(256x256x252)
|| notes
|| state 1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up
|| folder 0004
||

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented
||folder 0004
||(gucci1052-5-30-09-28406)

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up
||folder 0007
||

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented
||folder 0005
||

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact
||folder 0004
||

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented
||folder 0006
||very strong bias

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up
||folder 0006
||

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented
||folder 0005
||

|-
!ralph 06-08
|folder 0003 (first 252 slices)
||image not oriented
||
||

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented
||folder 0005
||

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
||folder 0005
||
|}
Note: almost all images have artifact under the skull.

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2010-01-05T17:53:05Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| baseline T1 data source(256x256x252)
|| notes
|| state 1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up
|| folder 0004
||

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented
||folder 0004
||(gucci1052-5-30-09-28406)

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up
||folder 0007
||

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented
||folder 0005
||

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact
||folder 0004
||

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented
||folder 0006
||very strong bias

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up
||folder 0006
||

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented
||folder 0005
||

|-
!ralph 06-08
|folder 0003 (first 252 slices)
||image not oriented

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented
||folder 0005
||

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
||folder 0005
||
|}
Note: almost all images have artifact under the skull.

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2010-01-05T17:34:28Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| baseline T1 data source(256x256x252)
|| notes
|| state 1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up
|| folder 0004
||

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented
||folder 0004
||(gucci1052-5-30-09-28406)

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!ralph 06-08
|folder 0003 (first 252 slices)
||image not oriented

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
|}
Note: almost all images have artifact under the skull.

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2010-01-05T17:32:44Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| baseline T1 data source(256x256x252)
|| notes
|| state 1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!ralph 06-08
|folder 0003 (first 252 slices)
||image not oriented

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
|}
Note: almost all images have artifact under the skull.

Proposed work pipeline

2009-10-12T14:30:29Z

Ginger:

# Ginger: produce ICC mask for chosen template: Gucci
# Andriy: use BRAINSfit to align all other sbjects to template through Affine + BSpline
# Ginger: compute average of the aligned subjects and manual separate to GM, WM + CSF
# Ron: manually segment sub-cortical structures.
# Ginger: produce Slicer3 module for back propagation and manual correct the results
# Ginger: smooth back propagation results to generate final templates
# Andriy: register atlas to each subjects and apply resulting transformation to align atlas to each subjects
# Ginger: finalize by EM segmentation.

Meanwhile: Ginger will perform EM segmentation on Gucci to separate into WM, GM, CSF and Ron helps to segment sub-cortical structures.

Proposed work pipeline

2009-10-12T14:30:17Z

Ginger:

# Ginger: produce ICC mask for chosen template: Gucci
# Andriy: use BRAINSfit to align all other sbjects to template through Affine + BSpline
# Ginger: compute average of the aligned subjects and manual separate to GM, WM + CSF
# Ron: manually segment sub-cortical structures.

# Ginger: produce Slicer3 module for back propagation and manual correct the results

# Ginger: smooth back propagation results to generate final templates

# Andriy: register atlas to each subjects and apply resulting transformation to align atlas to each subjects

# Ginger: finalize by EM segmentation.

Meanwhile: Ginger will perform EM segmentation on Gucci to separate into WM, GM, CSF and Ron helps to segment sub-cortical structures.

Proposed work pipeline

2009-10-12T14:29:58Z

Ginger:

# Ginger: produce ICC mask for chosen template: Gucci

# Andriy: use BRAINSfit to align all other sbjects to template through Affine + BSpline

# Ginger: compute average of the aligned subjects and manual separate to GM, WM + CSF

# Ron: manually segment sub-cortical structures.

# Ginger: produce Slicer3 module for back propagation and manual correct the results

# Ginger: smooth back propagation results to generate final templates

# Andriy: register atlas to each subjects and apply resulting transformation to align atlas to each subjects

# Ginger: finalize by EM segmentation.

Meanwhile: Ginger will perform EM segmentation on Gucci to separate into WM, GM, CSF and Ron helps to segment sub-cortical structures.

Proposed work pipeline

2009-10-12T14:29:39Z

Ginger:

# Ginger: produce ICC mask for chosen template: Gucci

#Andriy: use BRAINSfit to align all other sbjects to template through Affine + BSpline

#Ginger: compute average of the aligned subjects and manual separate to GM, WM + CSF

#Ron helps to manually segment sub-cortical structures.

#Ginger: produce Slicer3 module for back propagation and manual correct the results

#Ginger: smooth back propagation results to generate final templates

#Andriy: register atlas to each subjects and apply resulting transformation to align atlas to each subjects

#Ginger: finalize by EM segmentation.

Meanwhile: Ginger will perform EM segmentation on Gucci to separate into WM, GM, CSF and Ron helps to segment sub-cortical structures.

Proposed work pipeline

2009-10-12T14:28:41Z

Ginger: Created page with '# Ginger: produce ICC mask for chosen template: Gucci #Andriy: use BRAINSfit to align all other sbjects to template through Affine + BSpline #Ginger: compute average of the ali…'

# Ginger: produce ICC mask for chosen template: Gucci

#Andriy: use BRAINSfit to align all other sbjects to template through Affine + BSpline

#Ginger: compute average of the aligned subjects and manual separate to GM, WM + CSF

Ron helps to manually segment sub-cortical structures.

produce Slicer3 module for back propagation and manual correct the results

smooth back propagation results to generate final templates

#Andriy: register atlas to each subjects and apply resulting transformation to align atlas to each subjects

#Ginger: finalize by EM segmentation.

Meanwhile: Ginger will perform EM segmentation on Gucci to separate into WM, GM, CSF and Ron helps to segment sub-cortical structures.

NA-MIC NCBC Collaboration:Measuring Alcohol and Stress Interaction

2009-10-12T14:20:39Z

Ginger: /* Documentation */

Back to [[NA-MIC_External_Collaborations]]
==Abstract==
Alcohol abuse continues to be a major problem in the United States with an estimated 14 million adults
meeting the clinical criteria for a diagnosis of alcohol abuse or alcoholism at an estimated cost of $185
billion. Chronic alcohol exposure has been shown to impact structure and function of the brain. Since
most of what is known of alcohol's effects on the brain is based on studies of individuals who have
abused for a long time, it is unclear whether some of the effects might occur early in the abuse process.
Additionally, human studies are often complicated by many factors incuding polydrug abuse, poor
nutritional states and other medical conditions. Another complicating factor is the interaction of stress
and alcohol.

Using magnetic resonance imaging (MRI), we will examine the the effects of chronic alcohol self-
administration on brain structure and function in a group of monkeys that have been mother-reared or
nursery-reared. MRI scans will be acquired to determine whether there are differences in stucture or
functions as a result of alcohol exposure and whether rearing conditions play a role in these effects.
MRI techniques will also be used to examine how chronic alcohol consumption affects cerebral blood
flow, which is a measure of neuronal activity. Because we will conduct these measures both before
and after alcohol exposure in the same "young adult" animals, we will be able to track any changes that
might occur early in alcohol abuse which is something that cannot be accomplished in human studies.
Another goal is to adapt imaging tools and computer software that are used to measure the brains of
humans to the monkey brain. To accomplish this, we will collaborate with the National Alliance for
Medical Computing (NA-MIC), an existing NIH Center for Biomedical Computing.

We can control all variables associated with alcohol self-administration in these monkeys. They will be
monitored both before and after exposure to alcohol, thus, we will be able to monitor the progression of
any changes in structure or function that directly result from alcohol exposure. These studies will help
identify how exposure to long-term alcohol might affect how the brain works.

==Grant #==
1R01AA016748-01

==Key Personnel==
Wake Forest University: Jim Daunais, Bob Kraft

Virginia Tech: Chris Wyatt, Vidya Rajagopalan, Xiaoxing(Ginger) Li, Xiaojing Long

NAMIC: Kilian Pohl, Sandy Wells, Andriy Fedorov

==Projects==
Active projects in this collaboration are:

=== Tutorials on using EMSegmenter to segment nonhuman primate images ===
#Slicer3 based tutorial
##[http://www.na-mic.org/Wiki/index.php/Image:EMSegment_TrainingTutorial.pdf Tutorial Slide Handouts (PDF)]
##[http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Vervet Atlas]
##[http://www.na-mic.org/Wiki/images/c/c9/SlicerTutorialData_Vervet.tar.gz Example Data]
#[http://bsl.ece.vt.edu/download/NHP-EMSTut.zip older Slicer2.x based tutorial]
## [http://www.na-mic.org/Wiki/index.php/Image:EMSegment_Vervet_TrainingTutorial_Slicer2.pdf Vervet segmentation with Slicer2] (extracted from above for convenience)

=== Nonhuman Primate Atlas Development ===
Vervet Probabalistic Atlas - Adult
# [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Vervet Probabilistic Atlas (VPA-10)]

Rhesus Probabalistic Atlas - Adult
#[[Rhesus_WMGM_Atlas|WM/GM Atlas]]
#[[Rhesus_Subcortical_Atlas|Subcortical Atlas]]

Rhesus Probabalistic Atlas - Juvenile
#[[Rhesus_Juvenile_WMGM_Atlas|WM/GM Atlas]]

Cyno Probabalistic Atlas - Adult
#[[Cyno_WMGM_Atlas|WM/GM Atlas]]

=== Segmentation ===
#[[Rhesus_ICV_Extraction | Rhesus ICV Extraction]]
#[[Rhesus_EM_Segmentation | Rhesus EM Segmentation]]

=== Documentation ===
# [[Atlas-to-subject registration with FSL and Diffeomorphic Demons, for EM segmentation.]]
# [http://wiki.na-mic.org/Wiki/index.php/Vervet_MRI_registration Atlas-to-subject registration with BRAINSFit]
# [[On the performance of Slicer 2 and Slicer3 EM segmentation modules]]
# [[Proposed revised atlas construction workflow]]
# [[Current atlas construction workflow]]
# [[Proposed work pipeline]]

==Publications==
==Meetings and Events Specific to this Collaboration==
*[[2007_Programming/Project_Week_MIT|2007 Summer Programming/Project Week at MIT]]
*[[AHM_2008|2008 NA-MIC AHM]]
*[[NA-MIC/Projects/External Collaboration/Measuring Alcohol and Stress Interaction|2008 NA-MIC Summer Project Week Page]]
* 2009 NA-MIC Winter Project Week Pages: [[2009_Winter_Project_Week_WFU1|Project 1]]
* [http://wiki.na-mic.org/Wiki/index.php/Measuring_Alcohol_Stress_Interaction 2009 Summer project week Boston]
* [http://wiki.na-mic.org/Wiki/index.php/Events:VT-NAMIC-NCBC-Collaboration-meeting-2009 2009 October 7-9 VT-NAMIC collaboration meeting]

==Resource Links==

* [http://www.slicer.org slicer]
* [http://www.bsl.ece.vt.edu Bioimaging Systems Lab at VT]

Oct 08 2009 meeting: schedule for Oct 9

2009-10-09T02:27:44Z

Ginger: Created page with 'With Ron: * what caused the hole in Valentino data; * check the quality of N3 bias correction; * learn how to manually segment sub-cortical structures. With Sandy: * check t…'

With Ron:

* what caused the hole in Valentino data;

* check the quality of N3 bias correction;

* learn how to manually segment sub-cortical structures.

With Sandy:

* check the quality of the smoothed data;

With Chris:

* information about the stage1 data / data transfer issue;

With all:

* build up the list of non-slicer3 tools / scripts;

* discuss protocols on documentation;

* conform proposed work flow;

* discuss time-line / winter NAMIC meeting/ potential publications.

Events:VT-NAMIC-NCBC-Collaboration-meeting-2009

2009-10-09T02:19:49Z

Ginger: /* Progress */

== Logistics ==

* Andriy Fedorov and Ginger Li: work on pre-meeting agenda Wed-Thu October 7-8, 2009
* All participants meeting: Friday, '''October 9, 2009, 9am-4pm''' (time is tentative, but room has been reserved, confirmed with Katie Mastrogiacomo)
* Wednesday and Thursday location: [http://www.spl.harvard.edu/pages/Directions#Getting_to_1249_Boylston_Street. 1249 Boylston Street]
* Friday meeting location: [http://www.spl.harvard.edu/pages/Directions#Getting_to_1249_Boylston_Street. 1249 Boylston Street], 2nd floor conference room

== Goals ==

* To be completed or partially completed prior to the meeting by Andriy Fedorov and Ginger Li (Oct 7-8):
** Provide necessary user- and developer-level Slicer training for Ginger
** Understand the currently used processing workflow
** Confirm reproducibility of the processing steps between VT and SPL
** Document the processing steps as precisely as necessary for reproducibility and cross-validation of the results (wiki page)
* To be completed during the meeting
** Review and confirm the correctness of the workflow based on the wiki documentation (Chris, Ron and Sandy)
** Review the available data and current segmentation results
** Discuss action timeline for the remaining part of the funded project period
*** Potential for segmentation validation using cryo-histology (Chris can show some results of reconstructed 3D cryo-histo volumes)
*** Potential for cortical thickness measures / integration with Caret or freesurfer
** Discuss publication plans
** SLC meeting plans

== Agenda ==

*Review, discuss and finalize the processing workflow (to be discussed and documented by Andriy Fedorov and Ginger Li prior to the meeting Oct 7-8)
**atlas construction
**image pre-processing
**registration
**segmentation
*Review currents results
** Tommy and Louis images appear to have bad quality. Ron recommends not to use these subjects for atlas construction
{|
|<gallery perrow="3" heights="200px" widths="300px">
Image:oscar_snr.jpg|'''Oscar'''
Image:tommy_snr.jpg|'''Tommy'''
Image:louis_snr.jpg|'''Louis'''
</gallery>
|}
* Identify Slicer components of the workflow that are ... working, broken, missing

== Schedule ==
* Wednesday, Oct 7 plans:
** Walk through the segmentation workflow with EM Segmenter, verify scene saving, usability (Ginger)
** Document the current segmentation problems, prepare for discussion with Sandy
** 4pm: Meet with Sandy at Thorn to discuss current EM Segmenter results
** Install/verify installation of Slicer and related modules (BRAINSFit, N3MRILightCLI) (Andrey)
** Using and developing Slicer modules/extensions (Andrey)
** Using BRAINSFit (Andrey)
** AC-PC module usage for pose alignment (Ginger)
** Data transfer (images + segmentations)
** Aftenoon: Ron stops by to review progress (?)
* Friday, Oct 9, 9am (earlier?): Ron, Sandy, Ginger, Andrey meet at Boylston conference room to start discussion
* 10am: Chris' flight arrives at Logan
* 1pm: Chris gives a journal club talk at 1249 Boylston conference room
* 7pm: Chris' flight departs from Logan

== Progress ==

* Oct 7
** Reviewed available data at the initial time-point: [[October 07 2009 Meeting: VT-NAMIC Data review summary]]
** Discussed AC-PC usage module. Agreed possibly not helpful, since it does not align AC-PC line and midline between subjects, only straightens the AC-PC to be horizontal, and midline to be vertical. Manual alignment is needed prior to this to reliably locate AC and PC.
** Considering the data review (mis-alignment of the subjects in atlas; presense of bias; severe noise problems with some of the subject, while better data is available for the same subjects), and availability of the new tools (BRAINSFit robust up to BSplines, N3 bias correction in Slicer does good job -- can improve bias and registration) agreed that the atlas construction must be revisited
** Agreed on the conventions how to choose and load the data, all 10 subjects prepared in NRRD from the raw DICOM data, visually inspected for errors [[October 07 2009 Meeting: VT-NAMIC Data load protocol and comments]]
** Sandy suggested to include anisotropic diffusion into the atlas construction and segmentation workflow. EM Segmenter result can possibly be improved by adjusting the MFA parameter
** Ginger's laptop has problems running Slicer -- had to apply updates and are in the process of recompiling Slicer -- should be ready by Thursday
* Separate page for the [[VT-NAMIC Workflow Slicer-related issues]] -- to be revised and submitted as bug reports

* Oct 8
** Work on the revised atlas workflow: [[October 08 2009 Meeting: VT-NAMIC Revised atlas construction workflow]]
** To be discussed on Oct 9: [[Oct 08 2009 meeting: schedule for Oct 9]]

== Attendance ==

Open to all interested parties:

#Chris Wyatt
#Ginger Li
#Ron Kikinis
#Sandy Wells
#Andriy Fedorov

== References ==
* Measuring alcohol and stress interaction [http://wiki.na-mic.org/Wiki/index.php/NA-MIC_NCBC_Collaboration:Measuring_Alcohol_and_Stress_Interaction NCBC Collaboration page]
* [http://wiki.na-mic.org/Wiki/index.php/Atlas-to-subject_registration_with_FSL_and_Diffeomorphic_Demons,_for_EM_segmentation. Atlas construction workflow], put together by Ginger prior to the meeting
* [http://wiki.na-mic.org/Wiki/index.php/On_the_performance_of_Slicer_2_and_Slicer3_EM_segmentation_modules Slicer 2 vs. Slicer 3 EM Segmentation performance qualitative comparison], completed by Ginger prior to the meeting; also as a [http://www.na-mic.org/Wiki/index.php/Current_atlas_construction_workflow table]
* [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Atlas construction workflow] off BSL web site
* [http://www.na-mic.org/Wiki/index.php/Vervet_MRI_registration Evaluation of BRAINSFit performance on atlas-to-subject vervet registration], completed by Andrey prior to the meeting
* M. Styner, R. Knickmeyer, S. Joshi, C. Coe, S. J. Short, and J. Gilmore. Automatic brain segmentation in rhesus monkeys. Proc SPIE Medical Imaging Conference, Proc SPIE Vol 6512 Medical Imaging 2007, pp 65122L-1 - 65122L-8 [http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.cs.unc.edu%2F~styner%2Fpublic%2FMyPapers%2Fpapers%2FMedical%2520Imaging%25202007%2520Image%2520Processing%25202007%2520Styner-1.pdf&ei=DubDSt3AAciwlAfJp5TwBA&usg=AFQjCNGojUvBkh6TwbS0vJ9X7OEFLDFxqQ&sig2=wHgXTrdAXzUC5DuL_C7CCQ pdf]
* [http://wiki.na-mic.org/Wiki/index.php/Proposed_revised_atlas_construction_workflow Revised atlas construction workflow proposal], completed by Andriy Fedorov, communicated to Ron, Sandy, Sylvain Jaume, discussed with Sandy, all prior to the meeting

October 08 2009 Meeting: VT-NAMIC Revised atlas construction workflow

2009-10-09T02:17:42Z

Ginger:

== TODOs ==
* link the software tools and scripts with the workflow
* prepare a branch in the sandbox with the tools not available in Slicer (confirm with Chris copyright issues)

== Software ==

* Slicer/extensions versioning

== Workflow ==

# Prepare the data using the instructions -- LINK
# Gradient anisotropic diffusion -- (time step: 0.0625, iterations: 5)
# BRAINSFit initial registration -- PARAMETERS -- failure to recover the pose for Hugo and Valentino (both have severe head tilt in all three planes)
# Manual pose alignment for Hugo and Valentino
# Manual ICC for Tommy (Ginger, itkSNAP)
# Manual ICC label map smoothed using slicer smoothing label map model -( 10 iterations).
# Manual mask dilate for initial skull striping of other objects --(8-neighbors, 3 times).
# Resampling of the ICC to each subject
# Bias field correction with ICC
# Pipelined registration -- problems with ICC, resampling the transform

October 08 2009 Meeting: VT-NAMIC Revised atlas construction workflow

2009-10-09T02:17:25Z

Ginger:

== TODOs ==
* link the software tools and scripts with the workflow
* prepare a branch in the sandbox with the tools not available in Slicer (confirm with Chris copyright issues)

== Software ==

* Slicer/extensions versioning

== Workflow ==

# Prepare the data using the instructions -- LINK
# Gradient anisotropic diffusion -- (time step: 0.0625, iterations: 5)
# BRAINSFit initial registration -- PARAMETERS -- failure to recover the pose for Hugo and Valentino (both have severe head tilt in all three planes)
# Manual pose alignment for Hugo and Valentino
# Manual ICC for Tommy (Ginger, itkSNAP)
# Manual ICC label map smoothed using slicer smoothing label map model -( 10 iterations).
# Manual mask dilate for initial skull striping of other objects --(8-neighbours, 3 times).
# Resampling of the ICC to each subject
# Bias field correction with ICC
# Pipelined registration -- problems with ICC, resampling the transform

October 08 2009 Meeting: VT-NAMIC Revised atlas construction workflow

2009-10-09T02:14:48Z

Ginger:

== TODOs ==
* link the software tools and scripts with the workflow
* prepare a branch in the sandbox with the tools not available in Slicer (confirm with Chris copyright issues)

== Software ==

* Slicer/extensions versioning

== Workflow ==

# Prepare the data using the instructions -- LINK
# Gradient anisotropic diffusion -- (time step: 0.0625, iterations: 5)
# BRAINSFit initial registration -- PARAMETERS -- failure to recover the pose for Hugo and Valentino (both have severe head tilt in all three planes)
# Manual pose alignment for Hugo and Valentino
# Manual ICC for Tommy (Ginger, itkSNAP)
# Resampling of the ICC to each subject
# Bias field correction with ICC
# Pipelined registration -- problems with ICC, resampling the transform

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:38:28Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| T1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!ralph 06-08
|folder 0003 (first 252 slices)
||image not oriented

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
|}
Note: almost all images have artifact under the skull.

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:36:38Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| T1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!ralph 06-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
|}
Note: almost all images have artifact under the skull.

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:36:00Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
| T1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up

|-
!issac
|folder 0005 (first 252 slices)
|| data is noisy

image not oriented

|-
! Louis
|folder 0004 (first 252 slices)
||vessels light up

bad contrast in certain regions of the image

image not oriented

sever ringing artifact

|-
!marc
|folder 0003 (first 252 slices)
||noisy data

image not oriented

|-
! oscar
|folder 0003(first 252 slices)
||bias

image not oriented

a few vessels light up

|-
!ralph 04-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!ralph 06-08
|folder 0003 (first 252 slices)
|| vessels partially light up

image not oriented

|-
!tommy
|folder 0003 (first 252 slices)
|| data problem : bad contrast or noise

image not oriented

|-
!valentino
|folder 0004 (first 252 slices)
||vessels partially light up
|}
Note: almost all data has artifact under the skulls.

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:28:28Z

Ginger:

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:23:53Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
subject
| T1 data source(256x256x252)
|| notes

|-
! calvin
|Perform "parse directory", and select the first T1-SPGR sequence
||vessels light up

|-
!gucci
|folder 0004 (all 252 slices)
||vessels partially light up

image not oriented

|-
!hugo
|folder 0005 (first 252 slices)
||vessels partially light up

|}

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:23:13Z

Ginger:

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:19:52Z

Ginger:

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:18:12Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject
|| T1 data source(256x256x252)
|- notes

|- calvin
||Perform "parse directory", and select the first T1-SPGR sequence
|-vessels light up

|}

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:16:55Z

Ginger:

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|! subject !! T1 data source(256x256x252) !! notes
|! calvin!!Perform "parse directory", and select the first T1-SPGR sequence!!vessels light up

|}

October 07 2009 Meeting: VT-NAMIC Data load protocol and comments

2009-10-08T03:13:46Z

Ginger: Created page with '{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1" |- ! subject !! T1 data source !! notes |- ! calvin | || |}'

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! subject !! T1 data source !! notes
|-
! calvin
|

||

|}

Initial atlas construction workflow

2009-10-05T20:59:23Z

Ginger:

== Content ==

This page describes the current working pipeline of atlas construction for vervet data.

== Data ==

10 vervet subjects (2 subjects possibly have unacceptable image quality), T1 sequence

== Workflow / Tools ==

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! Workflow step !! Description !! Tools
|-
! 1. Preparation
|
1.1. Select the subject that is best oriented as the template image

1.2. Manually identify the ICC in the template subject.

1.3. Perform intensity calibration for all other subjects to the template.

1.4. 12 DOF registration of each of the subjects to the template.

1.5. Skull-strip each subject using dilated ICC of the template.

||
1.1. n/a

1.2. manual in SNAP.

1.3. Slicer Histogram Matching module

1.4. FSL-Flirt

1.5. Slicer Mask, Editor modules (verify the mask encloses actual ICC!)

|-
! 2. Atlas construction
|

2.1. Use the skullstriped image from Step 1.2 as the template to repeat Step 1.4.

2.2. Perform non-rigid alignment of the subjects affinely registered to the template.

2.3. Compute the atlas as the average.

||
2.1 see above

2.2. Diffeomorphic demons

2.3. Some tool in Slicer under development

|-
! 3. Probabilistic atlas construction
|

3.1. Segment WM/GM/CSF from the averaged template, manually edit to ensure accuracy

3.2. Slightly smooth the segmentations Styner et al. recommend kernel of 0.4 mm variance.

|

3.1. Kmeans + manual

3.2. Gaussian smoothing module in Slicer, 0.6 and 0.65 kernel size used for Valentino and Tommy, respectively.

*We did not perform back propagation as in Styner07 due to the large structural differences in the dataset.

|}

== References ==

# M. Styner, R. Knickmeyer, S. Joshi, C. Coe, S. J. Short, and J. Gilmore. Automatic brain segmentation in rhesus monkeys. Proc SPIE Medical Imaging Conference, Proc SPIE Vol 6512 Medical Imaging 2007, pp 65122L-1 - 65122L-8 [http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.cs.unc.edu%2F~styner%2Fpublic%2FMyPapers%2Fpapers%2FMedical%2520Imaging%25202007%2520Image%2520Processing%25202007%2520Styner-1.pdf&ei=DubDSt3AAciwlAfJp5TwBA&usg=AFQjCNGojUvBkh6TwbS0vJ9X7OEFLDFxqQ&sig2=wHgXTrdAXzUC5DuL_C7CCQ pdf]
# Balci S.K., Golland P., Wells W.M. Non-rigid Groupwise Registration using B-Spline Deformation Model. Insight Journal - 2007 MICCAI Open Science Workshop. [http://hdl.handle.net/1926/568 link]
# Previous descriptions of the atlas construction workflow: [http://wiki.na-mic.org/Wiki/index.php/Atlas-to-subject_registration_with_FSL_and_Diffeomorphic_Demons,_for_EM_segmentation. Summary by Ginger Li], [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Description off BSL atlas page]

Initial atlas construction workflow

2009-10-05T20:44:42Z

Ginger:

== Content ==

This page describes the current working pipeline of atlas construction for vervet data.

== Data ==

10 vervet subjects (2 subjects possibly have unacceptable image quality), T1 sequence

== Workflow / Tools ==

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! Workflow step !! Description !! Tools
|-
! 1. Preparation
|
1.1. Select the subject that is best oriented as the template image

1.2. Manually identify the ICC in the template subject.

1.3. Perform intensity calibration for all other subjects to the template.

1.4. 12 DOF registration of each of the subjects to the template.

1.5. Skull-strip each subject using dilated ICC of the template.

||
1.1. n/a

1.2. manual in SNAP.

1.3. Slicer Histogram Matching module

1.4. FSL/Flirt

1.5. Slicer Mask, Editor modules (verify the mask encloses actual ICC!)

|-
! 2. Atlas construction
|

2.1. Use the skullstriped image from Step 1.2 as the template to repeat Step 1.4.

2.2. Perform non-rigid alignment of the subjects affinely registered to the template.

2.3. Compute the atlas as the average.

||
2.1 see above

2.2. Diffeomorphic demons

2.3. Some tool in Slicer under development

|-
! 3. Probabilistic atlas construction
|

3.1. Segment WM/GM/CSF from the averaged template, manually edit to ensure accuracy

3.2. Slightly smooth the segmentations Styner et al. recommend kernel of 0.4 mm variance.

|

3.1. Kmeans + manual

3.2. Gaussian smoothing module in Slicer, 0.6 and 0.65 kernel size used for Valentino and Tommy, respectively.

*We did not perform back propagation as in Styner07 due to the large structural differences in the dataset.

|}

== References ==

# M. Styner, R. Knickmeyer, S. Joshi, C. Coe, S. J. Short, and J. Gilmore. Automatic brain segmentation in rhesus monkeys. Proc SPIE Medical Imaging Conference, Proc SPIE Vol 6512 Medical Imaging 2007, pp 65122L-1 - 65122L-8 [http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.cs.unc.edu%2F~styner%2Fpublic%2FMyPapers%2Fpapers%2FMedical%2520Imaging%25202007%2520Image%2520Processing%25202007%2520Styner-1.pdf&ei=DubDSt3AAciwlAfJp5TwBA&usg=AFQjCNGojUvBkh6TwbS0vJ9X7OEFLDFxqQ&sig2=wHgXTrdAXzUC5DuL_C7CCQ pdf]
# Balci S.K., Golland P., Wells W.M. Non-rigid Groupwise Registration using B-Spline Deformation Model. Insight Journal - 2007 MICCAI Open Science Workshop. [http://hdl.handle.net/1926/568 link]
# Previous descriptions of the atlas construction workflow: [http://wiki.na-mic.org/Wiki/index.php/Atlas-to-subject_registration_with_FSL_and_Diffeomorphic_Demons,_for_EM_segmentation. Summary by Ginger Li], [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Description off BSL atlas page]

Initial atlas construction workflow

2009-10-05T20:35:03Z

Ginger:

== Content ==

This page describes the current working pipeline of atlas construction for vervet data.

== Data ==

10 vervet subjects (2 subjects possibly have unacceptable image quality), T1 sequence

== Workflow / Tools ==

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! Workflow step !! Description !! Tools
|-
! 1. Preparation
|
1.1. Select the subject that is best oriented as the template image

1.2. Manually identify the ICC in the template subject.

1.3. Perform intensity calibration for all other subjects to the template.

1.4. 12 DOF registration of each of the subjects to the template.

1.5. Skull-strip each subject using dilated ICC of the template.

||
1.1. n/a

1.2. manual in SNAP.

1.3. Slicer Histogram Matching module

1.4. FSL/Flirt

1.5. Slicer Mask, Editor modules (verify the mask encloses actual ICC!)

|-
! 2. Atlas construction
|

2.1. Use the skullstriped image from Step 1.2 as the template to repeat Steps 1.3-1.4.

2.2. Perform non-rigid alignment of the subjects affinely registered to the template.

2.3. Compute the atlas as the average.

||
2.1 see above

2.2. Diffeomorphic demons

2.3. Some tool in Slicer under develop

|-
! 3. Probabilistic atlas construction
|

3.1. Segment WM/GM/CSF from the averaged template, manually edit to ensure accuracy

3.2. Slightly smooth the segmentations Styner et al. recommend kernel of 0.4 mm variance.

|

3.1. Kmeans + manual

3.2. Gaussian smoothing module in Slicer, 0.6 and 0.65 kernel size used for Valentino and Tommy, respectively.

*We did not perform back propagation due to the large structural differences in the dataset.

|}

== Standing questions ==

* when to perform bias correction? After Step 1.4?
* deep brain structures -- similar approach for identification in subjects as for GM/WM/CSF in Step 3.1?
* Styner et al. used T2 to segment CSF. We only have T1

== References ==

# M. Styner, R. Knickmeyer, S. Joshi, C. Coe, S. J. Short, and J. Gilmore. Automatic brain segmentation in rhesus monkeys. Proc SPIE Medical Imaging Conference, Proc SPIE Vol 6512 Medical Imaging 2007, pp 65122L-1 - 65122L-8 [http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.cs.unc.edu%2F~styner%2Fpublic%2FMyPapers%2Fpapers%2FMedical%2520Imaging%25202007%2520Image%2520Processing%25202007%2520Styner-1.pdf&ei=DubDSt3AAciwlAfJp5TwBA&usg=AFQjCNGojUvBkh6TwbS0vJ9X7OEFLDFxqQ&sig2=wHgXTrdAXzUC5DuL_C7CCQ pdf]
# Balci S.K., Golland P., Wells W.M. Non-rigid Groupwise Registration using B-Spline Deformation Model. Insight Journal - 2007 MICCAI Open Science Workshop. [http://hdl.handle.net/1926/568 link]
# Previous descriptions of the atlas construction workflow: [http://wiki.na-mic.org/Wiki/index.php/Atlas-to-subject_registration_with_FSL_and_Diffeomorphic_Demons,_for_EM_segmentation. Summary by Ginger Li], [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Description off BSL atlas page]

Initial atlas construction workflow

2009-10-05T20:34:37Z

Ginger:

== Content ==

This page describes the current working pipeline of atlas construction for vervet data.

== Data ==

10 vervet subjects (2 subjects possibly have unacceptable image quality), T1 sequence

== Workflow / Tools ==

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! Workflow step !! Description !! Tools
|-
! 1. Preparation
|
1.1. Select the subject that is best oriented as the template image

1.2. Manually identify the ICC in the template subject.

1.3. Perform intensity calibration for all other subjects to the template.

1.4. 12 DOF registration of each of the subjects to the template.

1.5. Skull-strip each subject using dilated ICC of the template.

||
1.1. n/a

1.2. manual in SNAP.

1.3. Slicer Histogram Matching module

1.4. FSL/Flirt

1.5. Slicer Mask, Editor modules (verify the mask encloses actual ICC!)

|-
! 2. Atlas construction
|

2.1. Use the skullstriped image from Step 1.2 as the template to repeat Steps 1.3-1.4.

2.2. Perform non-rigid alignment of the subjects affinely registered to the template.

2.3. Compute the atlas as the average.

||
2.1 see above

2.2. Diffeomorphic demons

2.3. Some tool in Slicer under develop

|-
! 3. Probabilistic atlas construction
|

3.1. Segment WM/GM/CSF from the averaged template, manually edit to ensure accuracy

3.2. Slightly smooth the segmentations Styner et al. recommend kernel of 0.4 mm variance.

|

3.1. Kmeans + manual

3.2. Gaussian smoothing module in Slicer, 0.6 and 0.65 kernel size used for Valentino and Tommy, respectively.

We did not perform back propagation due to the large structural differences in the dataset.

|}

== Standing questions ==

* when to perform bias correction? After Step 1.4?
* deep brain structures -- similar approach for identification in subjects as for GM/WM/CSF in Step 3.1?
* Styner et al. used T2 to segment CSF. We only have T1

== References ==

# M. Styner, R. Knickmeyer, S. Joshi, C. Coe, S. J. Short, and J. Gilmore. Automatic brain segmentation in rhesus monkeys. Proc SPIE Medical Imaging Conference, Proc SPIE Vol 6512 Medical Imaging 2007, pp 65122L-1 - 65122L-8 [http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.cs.unc.edu%2F~styner%2Fpublic%2FMyPapers%2Fpapers%2FMedical%2520Imaging%25202007%2520Image%2520Processing%25202007%2520Styner-1.pdf&ei=DubDSt3AAciwlAfJp5TwBA&usg=AFQjCNGojUvBkh6TwbS0vJ9X7OEFLDFxqQ&sig2=wHgXTrdAXzUC5DuL_C7CCQ pdf]
# Balci S.K., Golland P., Wells W.M. Non-rigid Groupwise Registration using B-Spline Deformation Model. Insight Journal - 2007 MICCAI Open Science Workshop. [http://hdl.handle.net/1926/568 link]
# Previous descriptions of the atlas construction workflow: [http://wiki.na-mic.org/Wiki/index.php/Atlas-to-subject_registration_with_FSL_and_Diffeomorphic_Demons,_for_EM_segmentation. Summary by Ginger Li], [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Description off BSL atlas page]

Initial atlas construction workflow

2009-10-05T20:24:25Z

Ginger:

== Content ==

This page describes the current working pipeline of atlas construction for vervet data.

== Data ==

10 vervet subjects (2 subjects possibly have unacceptable image quality), T1 sequence

== Workflow / Tools ==

{| class="wikitable" style="text-align:center; width:800px; height:200px" border="1"
|-
! Workflow step !! Description !! Tools
|-
! 1. Preparation
|
1.1. Select the subject that is best oriented as the template image

1.2. Manually identify the ICC in the template subject.

1.3. Perform intensity calibration to the template.

1.4. 12 DOF registration of each of the subjects to the template.

1.5. Skull-strip each subject using dilated ICC of the template.

||
1.1. n/a

1.2. manual in SNAP.

1.3. Slicer Histogram Matching module

1.4. FSL/Flirt

1.5. Slicer Mask, Editor modules (verify the mask encloses actual ICC!)

|-
! 2. Atlas construction
|

2.1. Use the affine registered image from Step 1.6 as the template to repeat Steps 1.1-1.6.

2.2. Perform non-rigid alignment of the subjects affinely registered to the template.

2.3. Compute the atlas as the average.

||
2.1 see above

2.2. Slicer BRAINSFit extension (BSpline), possibly Groupwise BSpline registration by Balci et al. [2]

2.3. Tool from step 1.6.

|-
! 3. Probabilistic atlas construction
|

3.1. Segment WM/GM/CSF from the averaged template, manually edit to ensure accuracy

3.2. Slightly smooth the segmentations Styner et al. recommend kernel of 0.4 mm variance.

3.3. Back-propagate WM/GM/CSF segmentations to the training subjects using the deformation field from Step 2.2.

3.4. Construct the probabilistic atlas by averaging of the tissue maps, normalize locally to 1, construct rejection class by inverting the sum of WM+GM+CSF.

|

3.1. Slicer editor module (thresholding, manual editing)

3.2. Gaussian smoothing module (check availability)

3.3. Slicer applyDeformation module (command line accessible only!)

3.4. Tool from Step 1.6, tool to be developed, Slicer Editor module

|}

== Standing questions ==

* when to perform bias correction? After Step 1.4?
* deep brain structures -- similar approach for identification in subjects as for GM/WM/CSF in Step 3.1?
* Styner et al. used T2 to segment CSF. We only have T1

== References ==

# M. Styner, R. Knickmeyer, S. Joshi, C. Coe, S. J. Short, and J. Gilmore. Automatic brain segmentation in rhesus monkeys. Proc SPIE Medical Imaging Conference, Proc SPIE Vol 6512 Medical Imaging 2007, pp 65122L-1 - 65122L-8 [http://www.google.com/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fwww.cs.unc.edu%2F~styner%2Fpublic%2FMyPapers%2Fpapers%2FMedical%2520Imaging%25202007%2520Image%2520Processing%25202007%2520Styner-1.pdf&ei=DubDSt3AAciwlAfJp5TwBA&usg=AFQjCNGojUvBkh6TwbS0vJ9X7OEFLDFxqQ&sig2=wHgXTrdAXzUC5DuL_C7CCQ pdf]
# Balci S.K., Golland P., Wells W.M. Non-rigid Groupwise Registration using B-Spline Deformation Model. Insight Journal - 2007 MICCAI Open Science Workshop. [http://hdl.handle.net/1926/568 link]
# Previous descriptions of the atlas construction workflow: [http://wiki.na-mic.org/Wiki/index.php/Atlas-to-subject_registration_with_FSL_and_Diffeomorphic_Demons,_for_EM_segmentation. Summary by Ginger Li], [http://www.bsl.ece.vt.edu/data/vervet_atlas/vervet.php Description off BSL atlas page]