Parrish Neuroimaging Group
  • Team »
    • Todd Parrish, PhD
    • Jennie Yufen Chen, PhD
    • Xue Wang, PhD
    • Kenneth Weber, DC, PhD
    • Xiaowei Song
    • Ellen Fitzmorris, B.S.
    • Suvarna Badhe
    • Brian Williams, BSN, RN
  • Projects »
    • Caffeine Research
    • Diffusion Tensor Imaging (DTI)
    • Temporal Signal to Noise in Time Series Data
    • Hemodynamic Response Function (HRF)
    • Transcranial Magnetic Stimulation (TMS)
    • Acupuncture
    • The Cerebrovascular Reserve (CVR)
    • Imaging of the Spinal Cord
    • Voxel Lesion Symptom Mapping (VLSM) in Chronic Stroke
  • fmri »
    • fMRI Basics Presentation
    • Clinical
  • Volunteer
  • Contact Us
  • Neuroimaging Course »
    • Neuroimaging Course »
      • Registration
    • Advanced Neuroimaging: Fundamentals and Methods
Home» Xiaowei Song’s Notes

Xiaowei Song’s Notes

Phantom QA

Human QA

  • fMRI
    • Motion related metrics
      • AFNI euclidean (mm)
      • FD and DVARS
    • niftitools’ QA (niftiqa)
    • fBIRN QA for EPI images: mirror of fBIRN QA metrics explanations.
  • sMRI
  • dMRI
    • Vanderbilt MASI DTI QA help page; mirror of the help page.

Robust pipelines:

  • ini default config files (resting-state, task)
  • sMRI (including snapshots for each step)
    1. Skull strip (ROBEX / BEaST/FSL BET/AFNI 3dSkullStrip/FreeSurfer stage 1-5)
    2. non-local means filtering (optional)
    3. Robust MRI registration (FreeSurfer)
    4. FSL 3-stages registration
      • dof-9
      • dof-12
      • FNIRT
    1. FAST segmentation
      • preparing for BBR registration
  • fMRI (check this page to see intermediate files’ meaning)

    1. dicom to nifti, automatically, extract information from header
    2. slice timing correction to 1st slice
    3. motion metrics scrubbing/censoring (FD, DVARS, 3dvolreg) and regressors generation (6-motion, FD, DVARS, Friston-24)
    4. CSF/WM nuiances PCA signals extration (default using first 6 PCs)
    5. BBR to anat, then normalized to template (one step to minimize distortion of fMRI signals)
    6. Optional smoothing
    7. ReHo calculation
    8. FC calculation and Fish’s Z transformation
  • dMRI

Linux misc:

  • Ubuntu recommanded since there is a NeuroDebian project, which will make neuroimaging life much easier
  • Viewing windows in a remote shell of linux
    • Easy way: Install MobaXterm
    • Advanced way: install PuTTy and Cygwin which I prefer.
  • With Virtualbox and PuTTy/MobaXterm, we can work seamlessly with a local linux server including viewing results of FreeSurfer/Caret. Just remember map 22 port from virtual-linux to windows.

QUEST misc

  • msubi, get a shell of a computing node
  • msubq, submit a command line to run in a computing node
  • qjobs, get my jobs’ status and working dir …
  • qnodeslive, get current working node to debug programs

NUNDA misc 

I wrote a command line version of XNAT database operation interface based on curl with XNAT API and pyxnat. I am trying to make using NUNDA easier.

  • xnatlogin  
  • xnatview.py -p project -q subject -s session
  • xnatget.py projID subjID sessID
  • xnatsess.py -r sessID
  • xnatput.py
NUNDA resources levels in a nutshell:
-  - projects
-      - subjects
-          - experiments
-              - scans
-              - reconstructions
-              - assessors

Always keep Backup in mind:

  • By duplicati  (Basic, easy to use): http://www.duplicati.com/
  • By duplicity (Advanced, easy to make script): 
    • http://duplicity.nongnu.org/
    • http://zertrin.org/projects/duplicity-backup/

Useful links:

  • Bash pitfalls for writing scripts



(c) 2014 Parrish Neuroimaging Group