Reference
7. Available recipes
Name: |
|
aarecipe_standard_tr1p1_ver00 |
Description: |
Standard TR=1.1s recipe |
|
Parameter set choice: |
||
Type |
Parameter set |
|
directory_conventions |
aap_directory_conventions_ver00 |
|
options |
aap_options_ver00 |
|
tasklist |
aap_tasklist_cbudefault_ver00 |
|
spmanalysis |
aap_spmanalysis_tr1p1_ver00 |
|
acq_details |
aap_acq_details_ver00 |
*** need sparse, ISS ***
7. Available parameter sets
To see the full list of settings in a parameter set, at the Matlab prompt type:
>> edit [parameter set name]
e.g.,
>> edit aap_directory_conventions_ver00 |
7.2. directory_conventions
Parameter set |
Description |
aap_directory_conventions_ver00 |
Example fields:BRSub-directories called “4d_files”, “proc_fieldmaps”, “raw_fieldmaps”, “structurals”BRRaw data in /cbu/wbic_dataBRCentral store for structurals /cbu/imagers/data/structuralsBRCentral store for skull-stripped structuralsBR/cbu/imagers/data/structurals/skull_strippedBRTemplate for normalisation /cbu/imagers/spm/spm99/templates/sbrain_avgBR152T1.imBR |
7.2. options
Parameter set |
Description |
aap_options_ver00 |
Example fields:BRProvide fairly verbose output while running script.BRAutomatically identify field maps and structurals in datasets.BRCopy structurals to central directory. |
7.2. tasklist
Parameter set |
Description |
aap_tasklist_cbudefault_ver00 |
Task list: BRaamod_newsubj_initBRaamod_autoidentifyseriesBRaamod_copyfieldmapsBRaamod_pvconvBRaamod_copystructuralBRaamod_ana4dto3dBRaamod_slicetimingBRaamod_realignBRaamod_undistBRaamod_coreg'BRaamod_normBRaamod_undist_reslice |
7.2. spmanalysis
Parameter set |
Description |
aap_spmanalysis_tr1p1_ver00 |
Example settings:BRTR=1.1; slicetime=0.0762; smoothing FWHM=10mmBR |
7.2. acq_details
Parameter set |
Description |
aap_spmanalysis_tr1p1_ver00 |
Settings mostly overridden by aap_userBR |
7. Individual parameters
Here is a list of the individual parameters. They are divided into sections, by their type (shown in bold). To fill a field, in section 5 of your user script, put:
aap.[type].field=[value];
For example,
aap.directory_conventions.rawdatadir=’/cbu/scratch2/rhodri.cusack/my_raw_data’;
BRAnchor(s7.3.)BR=== 7.3. acq_details === |
|
|
see section 4.4 for notes on filling these fields |
||
Field |
Typical value |
Description |
root |
'/cbu/scratch2/rhodri.cusack/newsimultaneous' |
Path to directory where processed study will be stored. This directory should exist. Must be assigned by user. |
subjects |
{'W030510.ls1' 'W030511.ls1' 'W030512.ls1'} |
List of subjects. Must be assinged by user. |
sessions |
{'ns1_block1','ns1_block2','ns2_passive','ns2_active'}; |
List of sessions. Must be assigned by user |
brukersessionnums |
{[4 5 8 7] [4 5 7 8] [4 5 8 7]} |
List of session numbers. Must be assigned by user. [usually done with aas_addsubject command, see] |
fieldmaps |
[9 11;9 11; 9 12] |
Session number of fieldmaps. Automatically detected by default |
newfieldmap |
[8; 7; 4] |
Automatically detected by default |
numdummies |
18 |
Number of dummy scans at the start of each session. Must be assigned by user. |
structurals |
{} |
Session number of structurals. Automatically detected by default |
Anchor(s7.3.)BR=== 7.3. directory_conventions === |
||
Field |
Typical value |
Description |
four_d_dir |
'4d_files' |
Name of subdirectory in subject directory for 4d analyze files from pvconv.pl |
proc_fieldmaps |
'proc_fieldmaps' |
Name of subdirectory in subject directory for processed fieldmaps |
raw_fieldmaps |
'raw_fieldmaps' |
Name of subdirectory in subject directory for raw fieldmaps |
structdirname |
'structurals' |
Name of subdirectory in subject directory for structurals |
rawdatadir |
'/cbu/imagers/wbic_data/' |
Where raw data are |
centralstore_structurals |
'/cbu/imagers/data/structurals' |
Where structurals should be put |
centralstore_ssstructurals |
'/cbu/imagers/data/structurals/skull_stripped' |
Where to find and put skull stripped structurals |
T1sstemplate |
/cbu/imagers/spm/spm99/templates/ sbrain_avg152T1.img |
Template file for normalisation |
subject_directory_format |
0: manually specify in aap.directory_conventions.subject_directory_namesBR 1: truncated subject names (e.g., W020123) {default}BR 2: automatically generated ordinal labels for subject directories (S01, S02, S03...)BRBR |
Format for subject output directories |
subject_filename_format |
0: manually specify in field aap.directory_conventions.subject_filenamesBR 1: truncated subject names (e.g., W020123) {default}BR 2: automatically generated ordinal labels for subject directories (S01, S02, S03...)BR |
Format for subject directories |
subject_directory_namesBR |
{‘myfirstsubjdir’,’blob’} |
Manually specified names of directories for subjects |
subject_filenames |
{‘myfirstsubjfilename’,’blob’} |
Manually specified names of EPI filenames for subjects |
Anchor(s7.3.)BR=== 7.3. options === |
|
|
Field |
Typical value |
Description |
verbose |
1 |
Give user regular updates of status when running tool |
copystructuraltocentralstore |
1 |
Copy structurals to central store |
deletestructuralaftercopyingtoc |
1 |
Delete local copy of structural (before skull stripping) after copying to central store |
autoidentifystructural |
1 |
Automatically find structurals in raw datasets |
autoidentifyfieldmaps |
1 |
Automatically find fieldmaps in raw datasets |
Anchor(s7.3.)BR=== 7.3. spmanalysis === |
|
|
Field |
Typical value |
Description |
TRs |
1.1 |
Repetition time |
slicetime |
0.0762 |
Time to acquire a single slice |
FWHM |
10 |
Smoothing kernel (mm) |
evoltime |
9.104 |
Evolution time of field maps |
structuralbetf |
0.5 |
Bet fractional threshold value to be used when skull stripping tb_ files |
Anchor(s7.3.)BR=== 7.3. tasklist === |
|
|
Field |
Typical value |
Description |
initialisationmodules |
{'aamod_checkparameters' 'aamod_make_subjects_short' 'aamod_autoidentifyseries'}BR |
List of initialisation modules, which are always run during processing |
stages |
{'aamod_newsubj_init', 'aamod_autoidentifyseries', 'aamod_copyfieldmaps', 'aamod_pvconv', 'aamod_copystructural', 'aamod_ana4dto3d', 'aamod_slicetiming', 'aamod_realign', 'aamod_undist', 'aamod_coreg', aamod_norm', 'aamod_undist_reslice'} |
List of modules to be called to perform processing. These are only run once each (i.e., if there is no file done_xxx) |
7. Initialisation modules
These modules are run every time aa_doprocessing is called, which is different from standard modules (see Error! Reference source not found.) which are only executed if they haven’t already been done (as recorded by a “done_” flag).
7.4. aamod_checkparameters.m
Check that no new fields have been created in the aap structure by the user to trap mis-typing.
7.4. aamod_make_subjects_short
EPI filenames are short versions of subject number – either session code (e.g., CBU030207) or ordinal labels (e.g., S03) if specified in directory_conventions. This module makes these shorter filenames.
aamod_evaluatesubjectnames
Evaluates any wildcards in the subject names.
7.4. aamod_autoidentifyseries.m
Identifies data series in the raw data that correspond to fieldmaps, MPRAGE structurals and realtime T maps. For speed, stored in .mat file in subject directory the first time, and loaded from this.
7. Modules
All of the work is done by a set of modules, each of which performs an individual processing stage. These modules are described below. Each will only be executed if they haven’t been done already, as signified by a “done_” flag (see Error! Reference source not found., Error! Reference source not found.)
7.5. aamod_newsubj_init.m
Makes the subject directory if it doesn’t exist already.
7.5. aamod_copyfieldmaps.m
Copies the raw fieldmaps to the specified “rawphase” directory. The data in this directory will be used by undistort to reconstruct the raw data into processed fieldmaps.
=== aamod_pvconv.m |
[no longer used] === |
Script to convert the raw Siemens DICOM data files (generally located in /mridata).
aamod_convertseries.m
Converts EPI data from Siemens DICOM to NIFTI format.
aamod_converttmaps.m
Converts realtime t maps into NIFTI format and puts them into folder realtime_tmaps in each subject directory
7.5. aamod_copystructural.m
Copies the structural to the central unit storage place for structurals (currently /cbu/imagers/data/structurals).
aamod_ana4dto3d.m [no longer used]
Converts the 4dfiles (output of pvconv) to many 3d files. Done on EPI sessions (as you are acquiring multiple volumes across time). You will have as many 3d files in your session as there are TRs (or volumes) in the session. Optional output includes the mean image, the mean variance image, and the variance slice. These are helpful diagnostic images. There is also the option to not write out the dummy scans (because they are simply there to let the scanner signal steady, and won't be included in your analysis).
7.5. aamod_slicetiming.m
Corrects for the staggered order of slice acquisition. This makes the data on each slice correspond to the same point in time. Otherwise the data on one slice could represent a point in time as far removed as 1/2 the TR from an adjacent slice. Generally one uses sinc interpolation.
7.5. aamod_realign.m
Causes all of the volumes in each session to be registered to the first volume in that session, then all volumes are registered to the first volume of the first session. So, for each session, you get a mat file with the parameters to register all the slices in that session to the first slice in that session. You also get an overall .mat file with all the parameters in relation to the first scan/first volume. This overall mat file is in your first session folder.
You will also get a text file of the realignment parameters--this will be used by SPM during model estimation to get rid of signal due to movement.
7.5. aamod_undist.m
Creates processed fieldmaps (placed in the processed fieldmaps directory), which are then used to correct distortions in the EPIs. This improves normalization and statistical power.
7.5. aamod_coreg.m
Puts the selected scans (usually the mean EPI image and the structural) in the same space. Usually one coregisters the structural to the mean EPI image (without reslicing), as this requires reorienting only one structural, as opposed to hundreds of EPIs.
7.5.1 aamod_norm.m
Transforms the scans into standardised space (given by a template image). One can skull-strip and normalize the structurals to a T1 template, then apply the parameters derived by the normalization to the EPIs. Or one can try masked EPI normalization, in which masks are created for areas of signal dropout in the EPI, and then the EPIs are normalized to an EPI template.
7.5.1 aamod_undist_reslice.m
Helpful script that writes out the realigned-undistorted-normalized-smoothed images (or just up to normalized, with no smoothing). Saves one from having to write out the EPI images for each of the aforementioned stages individually, by combining all these processes in one resampling step.
Helper functions
aas_addsubject
SYNTAX:
aap=aas_addsubject(aap,’subject file name’,[EPI series number 1, EPI series number 2,...]);
EXAMPLE
aap=aas_addsubject(aap,’*CBU050011/*’,[5 11]);
DESCRIPTION
This function adds the details of a subject to be analysed. The file name of the raw data and EPI session numbers are specified.
You may use filename wildcard characters (*) in the subject directory name.
||By default, the raw data is searched for in /mridata/cbu/ (add line like
aap.directory_conventions.rawdatadir='/imaging/rhodri/newsimultaneous';
to user script – see 4.3
The EPI series number should be that of the raw data series directories with names similar to
Series_004_CBU_EPI_Std_RealTime_30s
or
Series_002_CBU_EPI_Standard
aa_benchmark
Tells you how long each stage took to run (in seconds).