-
Notifications
You must be signed in to change notification settings - Fork 1
GE 3T EPI Pfile Conversion (mux2nii)
(typically 1-3 hours per subject) The resting state sequence from the scanner is custom built for us, and as such requires a number of extra steps. Manually downloading the Pfiles from the BIOTIC server "Lauterbur" one of the extra steps. Once the data is fully transferred, a second step is requied to 'recon' the raw Pfiles (conversion from frequency "K-space" with 32-channels and 3x multi-banding into image-space), and then saved into neurimaging data format NIFTI (.nii/.nii.gz).
There should be 3 sets of Pfiles, this is how they correspond to the scanner sequences:
Scanner | Time | Pfile | Description |
---|---|---|---|
EPI NoMUX Cal | 00:28 | ~700 mb, lowest numeric title | This is a scan with no multiband factor and minimal distortion, will later be used as an SBRef in the minimal preprocessing pipelines. |
EPI Bomap - Rev | 00:15 | ~300mb, higher numeric title | This is a few volumes of the RS sequence acquired in a reversed phase encode direction. Will be used to estimate and correct the distortion. |
EPI MUX 3MM RS | 08:06 | ~15gb, highest numeric title | This is the main resting state scan where we acquire 500 volumes in a sub-second TR. As you can see it is by far the largest, and if it is smaller than 15gb then something went wrong with the scan. |
$ ls -l /shared/uher/FORBOW/rawdata/100_C_20190320/RS/
-rwxrwxrwx 1 uher 217820581 Mar 20 2019 20190320-0630_E12985_P13824.7.bz2
-rw-r--r-- 1 uher 68132 Mar 20 2019 20190320-0630_E12985_P13824.7_hdr.txt
-rw-rw-rw- 1 uher 861 Mar 20 2019 20190320-0630_E12985_P13824.7_param.dat
-rw-r--r-- 1 uher 3703044 Mar 20 2019 20190320-0630_E12985_P13824.7_ref.h5
-rwxrwxrwx 1 uher 55872 Mar 20 2019 20190320-0630_E12985_P13824.7_vrgf.dat
-rwxrwxrwx 1 uher 122288141 Mar 20 2019 20190320-0630_E12985_P14336.7.bz2
-rw-r--r-- 1 uher 55172 Mar 20 2019 20190320-0630_E12985_P14336.7_hdr.txt
-rw-rw-rw- 1 uher 863 Mar 20 2019 20190320-0630_E12985_P14336.7_param.dat
-rw-r--r-- 1 uher 1236076 Mar 20 2019 20190320-0630_E12985_P14336.7_ref.h5
-rwxrwxrwx 1 uher 55872 Mar 20 2019 20190320-0630_E12985_P14336.7_vrgf.dat
-rwxrwxrwx 1 uher 3930176862 Mar 20 2019 20190320-0630_E12985_P14848.7.bz2
-rw-r--r-- 1 uher 55160 Mar 20 2019 20190320-0630_E12985_P14848.7_hdr.txt
-rw-rw-rw- 1 uher 861 Mar 20 2019 20190320-0630_E12985_P14848.7_param.dat
-rw-r--r-- 1 uher 1236076 Mar 20 2019 20190320-0630_E12985_P14848.7_ref.h5
-rwxrwxrwx 1 uher 55872 Mar 20 2019 20190320-0630_E12985_P14848.7_vrgf.dat
Note, each Pfile is compressed with bzip2 (showing .bz2
extension(), and is accompanied by _hdr.txt
, _vrgf.dat
_param.dat
and _ref.h5
files. Occasionally there may be extra sets of Pfiles, most likely from re-running the main sequence if excessive head motion was suspected. If there is a lone Pfile, lowest in numeric title and a small size, unaccompanied by the above dat files, this is likely a shim, and while unnecessary for analysis it may cause the mux2nii script to fail, so create a new folder in the subject RS directory called shim
and move that file there.
- Update: now this is automatically taken care of in the conversion script. The Pfile could also be EPI or spectroscopy (MRS) data from another study.
Once the rawdata
subject RS folder is organized, we will convert the RAW Pfiles into NIFTI, and compress the original raw Pfiles with bzip2 compression.
- Open Terminal
- Remote login to Jaylah:
ssh uher@jaylah
- Navigate to rawdata folder:
cd /shared/uher/FORBOW/rawdata/
- Run the script (in background, redirecting output to text file in subject directory):
nohup ./_1_scripts/FORBOW_1_run_mux2nii.sh 100_C >100_C_20190320/nohup_mux2nii_20190701-0420.txt 2>&1 &
- Monitor script with ps command:
ps aux | grep '100_C'
or htop command ('q' to quit htop) - Expect this script to take 1-2 hrs to complete on Jaylah (automatically, using 4-8 CPU Cores).
- When completed the participant's /RS/ folder will have both a compressed NIFTI
.nii.gz
and a Matlab.mat
output file for each Pfile. There will also be a copy of the Matlab conversion scriptmux2nii_PBIL.m
, the conversion output and error logfiles, respectively:mux2nii-20190701-0422.stdout
andmux2nii-20190701-0422.stderr
.
$ ls -l /shared/uher/FORBOW/rawdata/100_C_20190320/RS/
-rwxrwxrwx 1 uher 217820581 Mar 20 2019 20190320-0630_E12985_P13824.7.bz2
-rw-r--r-- 1 uher 68132 Mar 20 2019 20190320-0630_E12985_P13824.7_hdr.txt
-rw-r--r-- 1 uher 15951075 Jul 1 2019 20190320-0630_E12985_P13824.7.mat
-rw-r--r-- 1 uher 5645801 Jul 1 2019 20190320-0630_E12985_P13824.7.nii.gz
-rw-rw-rw- 1 uher 861 Mar 20 2019 20190320-0630_E12985_P13824.7_param.dat
-rw-r--r-- 1 uher 3703044 Mar 20 2019 20190320-0630_E12985_P13824.7_ref.h5
-rwxrwxrwx 1 uher 55872 Mar 20 2019 20190320-0630_E12985_P13824.7_vrgf.dat
-rwxrwxrwx 1 uher 122288141 Mar 20 2019 20190320-0630_E12985_P14336.7.bz2
-rw-r--r-- 1 uher 55172 Mar 20 2019 20190320-0630_E12985_P14336.7_hdr.txt
-rw-r--r-- 1 uher 12000878 Jul 1 2019 20190320-0630_E12985_P14336.7.mat
-rw-r--r-- 1 uher 3808284 Jul 1 2019 20190320-0630_E12985_P14336.7.nii.gz
-rw-rw-rw- 1 uher 863 Mar 20 2019 20190320-0630_E12985_P14336.7_param.dat
-rw-r--r-- 1 uher 1236076 Mar 20 2019 20190320-0630_E12985_P14336.7_ref.h5
-rwxrwxrwx 1 uher 55872 Mar 20 2019 20190320-0630_E12985_P14336.7_vrgf.dat
-rwxrwxrwx 1 uher 3930176862 Mar 20 2019 20190320-0630_E12985_P14848.7.bz2
-rw-r--r-- 1 uher 55160 Mar 20 2019 20190320-0630_E12985_P14848.7_hdr.txt
-rw-r--r-- 1 uher 1004385038 Jul 1 2019 20190320-0630_E12985_P14848.7.mat
-rw-r--r-- 1 uher 476552586 Jul 1 2019 20190320-0630_E12985_P14848.7.nii.gz
-rw-rw-rw- 1 uher 861 Mar 20 2019 20190320-0630_E12985_P14848.7_param.dat
-rw-r--r-- 1 uher 1236076 Mar 20 2019 20190320-0630_E12985_P14848.7_ref.h5
-rwxrwxrwx 1 uher 55872 Mar 20 2019 20190320-0630_E12985_P14848.7_vrgf.dat
-rw-r--r-- 1 uher forbow 0 Jul 1 2019 mux2nii-20190701-0422.stderr
-rw-r--r-- 1 uher forbow 34938 Jul 1 2019 mux2nii-20190701-0422.stdout
-rw-r--r-- 1 uher forbow 1721 Jul 1 2019 mux2nii_PBIL.m