TBSS (proyecto BIOFACE)

https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/TBSS/UserGuide

tbss_1_preproc

[osotolongo@detritus bioface]$ mkdir bftbss
[osotolongo@detritus bioface]$ cd bftbss/
[osotolongo@detritus bftbss]$ cp ../working/*FA* ./
[osotolongo@detritus bftbss]$ tbss_1_preproc *.nii.gz
processing 0002_dti_FA
processing 0003_dti_FA
processing 0004_dti_FA
processing 0005_dti_FA
processing 0006_dti_FA
...
...
...

tbss_2_reg

[osotolongo@detritus bftbss]$ tbss_2_reg -T
0002_dti_FA_FA_to_target
0003_dti_FA_FA_to_target
0004_dti_FA_FA_to_target
0005_dti_FA_FA_to_target
0006_dti_FA_FA_to_target
...
...
...

tbss_3_postreg

[osotolongo@detritus bftbss]$ tbss_3_postreg -T
using pre-chosen target for registration
transforming all FA images into MNI152 space
0002_dti_FA_FA
0003_dti_FA_FA
0004_dti_FA_FA
0005_dti_FA_FA
...
....
...

tbss_4_prestats

[osotolongo@detritus bftbss]$ tbss_4_prestats 0.2
creating skeleton mask using threshold 0.2
creating skeleton distancemap (for use in projection search)
projecting all FA data onto skeleton
 
now run stats - for example:  
randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design.mat -t design.con -n 500 --T2 -V
(after generating design.mat and design.con)

Test con lenguaje

[osotolongo@detritus bftbss]$ head bioface_data.csv 
PSubject,Subject,Edad,Sexo,Estudios,FACEmem
B001,0001,64,1,11,13
B002,0002,53,2,9,48
B003,0003,60,2,8,10
B004,0004,64,2,7,21
B005,0005,62,2,16,30
B006,0006,64,2,5,25
B007,0007,62,2,11,39
B008,0008,58,2,11,45
B009,0009,51,2,11,28
[osotolongo@detritus bftbss]$ grep -f subjects.list bioface_data.csv > sub_bioface_data.csv
[osotolongo@detritus bftbss]$ head sub_bioface_data.csv
B002,0002,53,2,9,48
B003,0003,60,2,8,10
B004,0004,64,2,7,21
B005,0005,62,2,16,30
B006,0006,64,2,5,25
B007,0007,62,2,11,39
B008,0008,58,2,11,45
B009,0009,51,2,11,28
B010,0010,62,2,6,17
B011,0011,54,2,16,67

Redo

[osotolongo@detritus bftbss]$ grep -v "`awk -F',' '{print $2}' sub_bioface_data.csv`" subjects.list 
0032

Borro el 0032 de subjects.list, y

[osotolongo@detritus bftbss]$ for x in `cat subjects.list`; do cp ../working/${x}_dti_FA.nii.gz ./; done
[osotolongo@detritus bftbss]$ awk -F"," '{print $3,$4,$5,$6}' sub_bioface_data.csv > comp_data.dat
[osotolongo@detritus bftbss]$ head comp_data.dat
53 2 9 48
60 2 8 10
64 2 7 21
62 2 16 30
64 2 5 25
62 2 11 39
58 2 11 45
51 2 11 28
62 2 6 17
54 2 16 67
[osotolongo@detritus bftbss]$ Text2Vest comp_data.dat design.mat
[osotolongo@detritus bftbss]$ head design.mat
/NumWaves 4
/NumPoints 83
/Matrix
53 2 9 48
60 2 8 10
64 2 7 21
62 2 16 30
64 2 5 25
62 2 11 39
58 2 11 45
[osotolongo@detritus bftbss]$ cat model.con
0 0 0 1
0 0 0 -1
[osotolongo@detritus bftbss]$ Text2Vest model.con design.con
[osotolongo@detritus bftbss]$ cat design.con
/NumWaves 4
/NumPoints 2
/Matrix
0 0 0 1
0 0 0 -1

y repito todos los pasos.

[osotolongo@detritus bftbss]$ tbss_1_preproc *.nii.gz
...
[osotolongo@detritus bftbss]$ tbss_2_reg -T
...
[osotolongo@detritus bftbss]$ tbss_3_postreg -T
...
[osotolongo@detritus bftbss]$ tbss_4_prestats 0.2
...

Statistics

ahora ejecuto el randomise,

[osotolongo@detritus bftbss]$ randomise -i stats/all_FA_skeletonised.nii.gz -o bftbss -m stats/mean_FA_skeleton_mask -d design.mat -t design.con -n 1000 --T2 -V
randomise options: -i stats/all_FA_skeletonised.nii.gz -o bftbss -m stats/mean_FA_skeleton_mask -d design.mat -t design.con -n 1000 --T2 -V 
Loading Data: 
Data loaded
4.97831e+111 permutations required for exhaustive test of t-test 1
Doing 1000 random permutations
Starting permutation 1 (Unpermuted data)
Starting permutation 2
Starting permutation 3
Starting permutation 4
Starting permutation 5
...
...
...

Vamos a mirarlo,

[fsluser@FSLVm7_64 bftbss]$ fslview_deprecated /usr/local/fsl/data/standard/MNI152_T1_1mm stats/mean_FA_skeleton -l Green -b 0.2,0.7 bftbss_tfce_corrp_tstat1 -l Red-Yellow -b 0.95,1

y hacemos un tbss_fill,

[osotolongo@detritus bftbss]$ tbss_fill bftbss_tfce_corrp_tstat1 0.95 stats/mean_FA tbss_fill

para verlo mejor,

[fsluser@FSLVm7_64 bftbss]$ fslview_deprecated stats/mean_FA stats/mean_FA_skeleton -l Green -b 0.2,0.8 tbss_fill -l Red-Yellow -b 0.05,0.2

TBSS vs NP Composites

Lo primero es que tengo que tener los composites de NP hechos, esto de aqui: Sacando los Composite Scores de NP para BIOFACE.

Ahora he de unir estos composites con las variables demograficas que actuaran como covariables. En este caso, edad, sexo y educacion. Vamos a ver lo que tengo,

[osotolongo@brick03 tbss]$ cp ../ccs/bioface_composite_scores.csv ./
 
[osotolongo@brick03 tbss]$ head bioface_composite_scores.csv 
,Attention.and.Working.Memory,Processing.speed.and.Executive.function,Language,Verbal.Learning.and.Memory,Praxis,Visual.gnosis
B072,0.940383261534547,0.712149416119579,0.325607497357412,-0.550959332389243,0.421836370550401,0.445705493949369
B060,-0.200675612140164,0.605014438691104,0.325607497357412,0.546888909234256,-3.77634938935993,-0.906684675087603
B041,-1.34173448581487,-2.05628737448506,-1.98722088956859,-0.398317187495001,-1.64011037957474,-0.56858713282836
B051,0.374478120167295,0.963714319556467,0.325607497357412,-0.242472701174079,0.421836370550401,-1.13138977880836
B038,-0.191427021199958,0.588076018557911,0.325607497357412,-0.751279560842757,0.436694822482413,0.219960503550505
B052,1.51553699384201,0.966079684763052,0.0209050508660859,-0.167752799440297,0.421836370550401,0.670410597670127
B026,0.940383261534547,0.624957199529257,-1.07311355009461,-1.77280673001914,0.421836370550401,0.276136779480695
B040,0.940383261534547,0.716084275587973,0.325607497357412,0.429294436081222,0.421836370550401,0.670410597670127
B006,-1.34173448581487,0.416707414517106,0.0209050508660859,0.665373444688287,-1.68468573537077,0.0998024379954218
 
[osotolongo@brick03 tbss]$ head bioface_demog.csv 
Subject,Edad_visita,Sexo,Anyos_escolaridad
B072,66,0,7
B060,59,1,12
B041,63,1,11
B051,61,1,9
B038,63,1,10
B052,62,1,9
B026,62,1,20
B040,62,1,17
B006,64,1,6

Uf! a editar un poco,

[osotolongo@brick03 tbss]$ sed -i 's/^,/PSubject,/' bioface_composite_scores.csv
[osotolongo@brick03 tbss]$ sed -i 's/Subject/PSubject/;s/Edad_visita/Age/;s/Sexo/Gender/;s/Anyos_escolaridad/Education/' bioface_demog.csv 
[osotolongo@brick03 tbss]$ (head -n 1 bioface_composite_scores.csv && tail -n +2 bioface_composite_scores.csv | sort -t, -k 1) > bioface_composite_scores_sorted.csv 
[osotolongo@brick03 tbss]$ (head -n 1 bioface_demog.csv && tail -n +2 bioface_demog.csv | sort -t, -k 1) > bioface_demog_sorted.csv
[osotolongo@brick03 tbss]$ join -t, bioface_demog_sorted.csv bioface_composite_scores_sorted.csv > bioface_np_data.csv

y ahora me boy a buscar los codigos de neuroimagen,

[osotolongo@brick03 tbss]$ sed 's/;/,/g' ../bioface_mri.csv | sort -t, -k 2 > bioface_ni.csv 
[osotolongo@brick03 tbss]$ sed 's/;/,/g' ../bioface_mri.csv | sort -t, -k 2 | sed '1iSubject,PSubject' > bioface_ni.csv 
[osotolongo@brick03 tbss]$ join -t, -1 2 -2 1 bioface_ni.csv bioface_np_data.csv > bioface_data.csv
[osotolongo@brick03 tbss]$ (head -n 1 bioface_data.csv && tail -n +2 bioface_data.csv | sort -t, -k 2) > bioface_data_sorted.csv

y me quedan los datos juntos y ordenados por el id de neuroimagen,

[osotolongo@brick03 tbss]$ head bioface_data_sorted.csv
PSubject,Subject,Age,Gender,Education,Attention.and.Working.Memory,Processing.speed.and.Executive.function,Language,Verbal.Learning.and.Memory,Praxis,Visual.gnosis
B001,0001,64,0,13,0.940383261534547,0.258400126101346,0.325607497357412,0.351372192920761,0.421836370550401,0.437899980254665
B002,0002,53,1,10,-0.766580753507416,-0.640726365145137,0.325607497357412,-0.506483590256652,-1.65496883150675,-0.680939684688739
B003,0003,60,1,8,-0.775829344447622,-1.92758435997598,-0.588499842116566,0.188411915334138,0.451553274414425,-0.62476340875855
B004,0004,64,1,7,0.365229529227089,-0.078687424137654,-0.28379739562524,-0.313279153068839,0.421836370550401,-0.906684675087603
B005,0005,62,1,16,0.940383261534547,0.336724978845789,0.325607497357412,-0.205112750307188,0.421836370550401,-0.906684675087603
B006,0006,64,1,6,-1.34173448581487,0.416707414517106,0.0209050508660859,0.665373444688287,-1.68468573537077,0.0998024379954218
B007,0007,62,1,9,-1.34173448581487,-1.36008617577211,-1.50260718159054,-0.230553397768359,0.421836370550401,-3.05438810752004
B008,0008,58,1,15,-0.200675612140164,1.39604181164675,0.325607497357412,0.112090842887017,0.421836370550401,1.00850813992937
B009,0009,51,1,9,0.374478120167295,0.0671339898083944,0.325607497357412,0.465053216234773,-1.68468573537077,0.670410597670127

Lo que necesito ahora es quitar de esta lista todos los que no tengan DTI preprocesado,

[osotolongo@brick03 tbss]$ ls data
0002_dti_FA.nii.gz  0017_dti_FA.nii.gz  0036_dti_FA.nii.gz  0051_dti_FA.nii.gz  0070_dti_FA.nii.gz  0085_dti_FA.nii.gz
0003_dti_FA.nii.gz  0018_dti_FA.nii.gz  0037_dti_FA.nii.gz  0052_dti_FA.nii.gz  0071_dti_FA.nii.gz  0086_dti_FA.nii.gz
0004_dti_FA.nii.gz  0019_dti_FA.nii.gz  0038_dti_FA.nii.gz  0054_dti_FA.nii.gz  0072_dti_FA.nii.gz  0087_dti_FA.nii.gz
0005_dti_FA.nii.gz  0020_dti_FA.nii.gz  0039_dti_FA.nii.gz  0055_dti_FA.nii.gz  0073_dti_FA.nii.gz  0088_dti_FA.nii.gz
0006_dti_FA.nii.gz  0021_dti_FA.nii.gz  0040_dti_FA.nii.gz  0056_dti_FA.nii.gz  0074_dti_FA.nii.gz  0089_dti_FA.nii.gz
0007_dti_FA.nii.gz  0022_dti_FA.nii.gz  0041_dti_FA.nii.gz  0058_dti_FA.nii.gz  0075_dti_FA.nii.gz  0090_dti_FA.nii.gz
0008_dti_FA.nii.gz  0023_dti_FA.nii.gz  0042_dti_FA.nii.gz  0059_dti_FA.nii.gz  0076_dti_FA.nii.gz  0091_dti_FA.nii.gz
0009_dti_FA.nii.gz  0024_dti_FA.nii.gz  0043_dti_FA.nii.gz  0061_dti_FA.nii.gz  0077_dti_FA.nii.gz  0092_dti_FA.nii.gz
0010_dti_FA.nii.gz  0027_dti_FA.nii.gz  0044_dti_FA.nii.gz  0062_dti_FA.nii.gz  0078_dti_FA.nii.gz  0093_dti_FA.nii.gz
0011_dti_FA.nii.gz  0028_dti_FA.nii.gz  0045_dti_FA.nii.gz  0064_dti_FA.nii.gz  0079_dti_FA.nii.gz  0094_dti_FA.nii.gz
0012_dti_FA.nii.gz  0029_dti_FA.nii.gz  0046_dti_FA.nii.gz  0065_dti_FA.nii.gz  0080_dti_FA.nii.gz  0095_dti_FA.nii.gz
0013_dti_FA.nii.gz  0030_dti_FA.nii.gz  0047_dti_FA.nii.gz  0066_dti_FA.nii.gz  0081_dti_FA.nii.gz  0096_dti_FA.nii.gz
0014_dti_FA.nii.gz  0031_dti_FA.nii.gz  0048_dti_FA.nii.gz  0067_dti_FA.nii.gz  0082_dti_FA.nii.gz
0015_dti_FA.nii.gz  0034_dti_FA.nii.gz  0049_dti_FA.nii.gz  0068_dti_FA.nii.gz  0083_dti_FA.nii.gz
0016_dti_FA.nii.gz  0035_dti_FA.nii.gz  0050_dti_FA.nii.gz  0069_dti_FA.nii.gz  0084_dti_FA.nii.gz

asi que saco la lista, y tomo solo los datos de estos sujetos,

[osotolongo@brick03 tbss]$ ls data | awk -F"_" '{print $1}' > fas.list 
[osotolongo@brick03 tbss]$ (head -n 1 bioface_data_sorted.csv && while read -r x; do awk -v sbj="${x}" -F"," '{if($2==sbj) print}' bioface_data_sorted.csv; done <  fas.list) > bioface_data_dti.csv
[osotolongo@brick03 tbss]$ head bioface_data_dti.csv
PSubject,Subject,Age,Gender,Education,Attention.and.Working.Memory,Processing.speed.and.Executive.function,Language,Verbal.Learning.and.Memory,Praxis,Visual.gnosis
B002,0002,53,1,10,-0.766580753507416,-0.640726365145137,0.325607497357412,-0.506483590256652,-1.65496883150675,-0.680939684688739
B003,0003,60,1,8,-0.775829344447622,-1.92758435997598,-0.588499842116566,0.188411915334138,0.451553274414425,-0.62476340875855
B004,0004,64,1,7,0.365229529227089,-0.078687424137654,-0.28379739562524,-0.313279153068839,0.421836370550401,-0.906684675087603
B005,0005,62,1,16,0.940383261534547,0.336724978845789,0.325607497357412,-0.205112750307188,0.421836370550401,-0.906684675087603
B006,0006,64,1,6,-1.34173448581487,0.416707414517106,0.0209050508660859,0.665373444688287,-1.68468573537077,0.0998024379954218
B007,0007,62,1,9,-1.34173448581487,-1.36008617577211,-1.50260718159054,-0.230553397768359,0.421836370550401,-3.05438810752004
B008,0008,58,1,15,-0.200675612140164,1.39604181164675,0.325607497357412,0.112090842887017,0.421836370550401,1.00850813992937
B009,0009,51,1,9,0.374478120167295,0.0671339898083944,0.325607497357412,0.465053216234773,-1.68468573537077,0.670410597670127
B010,0010,62,1,8,-1.34173448581487,0.580441125865297,-0.28379739562524,-0.161348116586939,0.421836370550401,-0.906684675087603

Y ahora para cada variable a analizar solo tengo que construir el design.mat y ejecutar el randomise apropiado. A ver que datos tengo,

[osotolongo@brick03 tbss]$ head -n 1 bioface_data_sorted.csv | sed 's/,/\n/g' | cat -n
     1	PSubject
     2	Subject
     3	Age
     4	Gender
     5	Education
     6	Attention.and.Working.Memory
     7	Processing.speed.and.Executive.function
     8	Executive.function.verbal
     9	Language
    10	Verbal.Learning.and.Memory
    11	Praxis
    12	Visual.gnosis

Por otra parte para los DTI voy a hacer,

[osotolongo@brick04 data]$ tbss_1_preproc *.nii.gz && tbss_2_reg -T && tbss_3_postreg -T && tbss_4_prestats 0.2

OJO: Esto demora bastante, asi que lo envias y ahi lo dejas, con suerte ya mañana estara hecho!

y despues de todo esto lo que hacemos es lanzar los randomise a traves de slurm,

prandomise.py

#!/usr/bin/python3
 
"""
This program executes FSL randomise command over a given set of external
variables. You should supply the FSL template for the group of subjects,
the variables list, the full curated data and the design matrices. The 
script only  parallelize the set of randomise calls over the cluster.
 
"""
 
import os
import sys
import getopt
import csv
import re
import tempfile
import subprocess
 
# Static covariables
stcov = [3, 4, 5]
# Variables a procesar
dynvars = list(range(6,13))
 
#Entorno de trabajo
time = '12:0:0'
cpus = 8
wdir = os.environ.get('PWD')
outdir = wdir+'/slurm'
datadir = wdir+'/data'
if not os.path.isdir(outdir): os.mkdir(outdir)
 
# Get CLI inputs
short_args = 'd:t:'
long_args = ['data=', 'template=']
 
try:
        args, values = getopt.getopt(sys.argv[1:], short_args, long_args)
except getopt.error as err:
        print(str(err))
        sys.exit(2)
 
for a,v in args:
        if a in ('--data', '-d'):
                datafile = v
        elif a in ('--template', '-t'):
                templatefile = v
try:        
        templatefile
except NameError:
        templatefile = datadir + '/stats/all_FA_skeletonised.nii.gz'
 
for v in dynvars:
        tdir = wdir + '/tbss_cs' + str(v)
        if not os.path.isdir(tdir): os.mkdir(tdir)
        ltw = ''
        with open(datafile) as f:
                reader = csv.reader(f)
                for row in reader:
                        if not row[0] == 'PSubject':
                                for sv in stcov:
                                        ltw += row[sv-1]+' '
                                ltw += row[v-1]+'\n'
        f.close()
        tempfile = 'tempdata_'+str(v)+'.dat'
        df = open(tempfile, 'w')
        df.write(ltw)
        df.close()
        os.system('Text2Vest '+df.name+' '+tdir+'/design.mat')
        os.system('Text2Vest model.con '+tdir+'/design.con')
for cs in dynvars:
        tdir = wdir + '/tbss_cs' + str(cs)
        bcontent = '#!/bin/bash\n'+'#SBATCH -J randomise\n'+'#SBATCH -c '+str(cpus)+'\n'+'#SBATCH --mem-per-cpu=4G\n'
        bcontent += '#SBATCH --time='+time+'\n'+'#SBATCH --mail-type=FAIL,TIME_LIMIT,STAGE_OUT\n'
        bcontent += '#SBATCH --mail-user='+os.environ.get('USER')+'\n'+'#SBATCH -o '+outdir+'/randomise-%j'+'\n'
        bcontent += 'randomise -i '+templatefile+' -o '+tdir+'/tbss_cs'+str(cs)+' -m '+datadir+'/stats/mean_FA_skeleton_mask -d '
        bcontent += tdir+'/design.mat -t '+tdir+'/design.con -n 5000 --T2 -V\n'
        sbfilename = outdir+'/run_cs_'+str(cs)+'.sh'
        sbf = open(sbfilename, 'w')
        sbf.write(bcontent)
        sbf.close()
        jobid = subprocess.getoutput('sbatch --parsable '+sbfilename)
        bcontent = '#!/bin/bash\n'+'#SBATCH -J randomise\n'+'#SBATCH -c '+str(cpus)+'\n'+'#SBATCH --mem-per-cpu=4G\n'
        bcontent += '#SBATCH --time='+time+'\n'+'#SBATCH --mail-type=FAIL,TIME_LIMIT,STAGE_OUT\n'
        bcontent += '#SBATCH --mail-user='+os.environ.get('USER')+'\n'+'#SBATCH -o '+outdir+'/cluster-%j'+'\n'
        bcontent += 'cluster --in='+tdir+'/tbss_cs'+str(cs)+'_tfce_corrp_tstat1 --thresh=0.95 -o '+tdir+'/clusters_cs'+str(cs)+' > '+tdir+'/clusters_cs'+str(cs)+'_index.txt\n'
        sbfilename = outdir+'/cluster_cs_'+str(cs)+'.sh'
        sbf = open(sbfilename, 'w')
        sbf.write(bcontent)
        sbf.close()
        os.system('sbatch --dependency=afterok:'+jobid+' '+sbfilename)
bcontent = '#!/bin/bash\n'+'#SBATCH -J randomise\n'
bcontent += '#SBATCH --mail-type=END\n'
bcontent += '#SBATCH --mail-user='+os.environ.get('USER')+'\n'+'#SBATCH -o '+outdir+'/randomise-%j'+'\n'
bcontent += ':\n'
sbfilename = outdir+'/run_cs_end.sh'
sbf = open(sbfilename, 'w')
sbf.write(bcontent)
sbf.close()
os.system('sbatch --dependency=singleton '+sbfilename)

y al lanzarlo, crea el entorno adecuado, ejecuta los randomise, y extrae los resultados posteriormente,

[osotolongo@brick04 tbss]$ ./prandomise.py -d bioface_data_dti.csv 
[osotolongo@brick04 tbss]$ head tbss_cs6/design.mat 
/NumWaves 4
/NumPoints 87
/Matrix
53 1 10 -0.766580753507416
60 1 8 -0.775829344447622
64 1 7 0.365229529227089
62 1 16 0.940383261534547
64 1 6 -1.34173448581487
62 1 9 -1.34173448581487
58 1 15 -0.200675612140164
 
[osotolongo@brick04 tbss]$ queue
   JOBID PARTITION                                 NAME     USER ST       TIME  NODES NODELIST(REASON)
  218903      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218905      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218907      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218909      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218911      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218913      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218914      fast                            randomise osotolon PD       0:00      1 (Dependency)
  218902      fast                            randomise osotolon  R      42:18      1 brick01
  218904      fast                            randomise osotolon  R      42:18      1 brick01
  218906      fast                            randomise osotolon  R      42:18      1 brick01
  218908      fast                            randomise osotolon  R      42:18      1 brick01
  218910      fast                            randomise osotolon  R      42:18      1 brick01
  218912      fast                            randomise osotolon  R      42:18      1 brick01

y ahi tienes para tres horas mas de calculo, pero despues(llega email de finalizacion) podemos saber rapidamente done tenemos resultados haciendo,

[osotolongo@brick03 tbss]$ for x in $(seq 6 12); do echo " var ${x} "; cat tbss_cs${x}/*_index.txt; done
 var 6 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)
 var 7 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)
 var 8 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)
 var 9 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)
1	53880	0.994	100	144	93	89	110	88.1
 var 10 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)
1	55983	0.997	58	59	80	89.4	107	87.4
 var 11 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)
 var 12 
Cluster Index	Voxels	MAX	MAX X (vox)	MAX Y (vox)	MAX Z (vox)	COG X (vox)	COG Y (vox)	COG Z (vox)

Ahora hago un tbss_fill de los que tienen resultados,

[osotolongo@brick03 tbss]$ tbss_fill tbss_cs9/tbss_cs9_tfce_corrp_tstat1.nii.gz 0.95 data/stats/mean_FA tbss_cs9/tbss_fill
[osotolongo@brick03 tbss]$ tbss_fill tbss_cs10/tbss_cs10_tfce_corrp_tstat1.nii.gz 0.95 data/stats/mean_FA tbss_cs10/tbss_fill

y los miro encima de la plantilla,

[osotolongo@brick03 tbss]$ fslview_deprecated data/stats/mean_FA data/stats/mean_FA_skeleton -l Green -b 0.2,0.8 tbss_cs9/tbss_fill -l Red-Yellow -b 0.05,0.2

Language	Verbal Learning and Memory

Detritus Wiki

Table of Contents