1 of 23

Neuroimaging Tools

Dr. Erin Dickie & Sophie Lafaille

2 of 23

Scientific Computing Workgroup

Have questions?

​

Office Hours Tuesdays 1-2

College Site Room 163

​

* Have ideas on improving scientific computing @ CAMH? Please join the group, we’d love to have you!

3 of 23

What are we gonna talk about?

  1. Type of Scans
  2. MR File Formats
  3. Analysis of Brain Anatomy
  4. Analysis of fMRI
  5. Analysis of DTI
  6. Viewing MR Images
  7. Pipelines Tools - Putting it all together

4 of 23

Types of Scans

T1-weighted

T2-weighted

fMRI

DTI

PET

5 of 23

File Formats

dcm2nii can convert DICOM files to NIFTI & NIFTI compressed (.gz) format

  • it also creates a “reoriented” and a “cropped” version of the file

“Raw” Scanner File Format

Example: DICOM

Analysis File Format

Example: NIfTI (Neuroimaging Informatics Technology Initiative)

​

6 of 23

Analysis of Brain Anatomy

what do you want to measure?

Malpass, Nature Reviews Neurology, 2011

Maguire et al, PNAS, 2000

FSL- FIRST analysis

7 of 23

Analysis for Brain Anatomy

Pipitone et al Neuroimage 2014

Freesurfer’s recon-all pipeline

MNI’s CIVET

MAGeT

Brain

Segmentation

8 of 23

Analysis of DTI - what to measure?

Cell membranes (and myelin) limit water diffusion

DTI measures the diffusion of water to infer tissue microstructure

9 of 23

Analysis of DTI - where (in the brain to measure from?)

Voineskos AN, et al. JAMA Psychiatry. 2013

Ameis SH, Am J Psychiatry. 2016

Tract Based Spatial Statistics (TBSS)

Tractography

10 of 23

What does fMRI measure??

Blood Oxygen Level Dependant (BOLD) signal

Blood Oxygen Level Dependant (BOLD) signal

11 of 23

Analysis of functional MRI- what can we measure?

​

Task bask Activation

  • Event related design
  • Block related design

​

Functional connectivity “resting state”

  • Independent components analysis
  • Seed based analysis

Matthews & Jezzard, Neurology, Neurosurgery & Psychiatry, 2004

12 of 23

Analysis of functional MRI - the packages

​

​

FMRIB Software Library (FSL)

Analysis Group, FMRIB, Oxford, UK

​

​

​

​

​

AFNI (Analysis of Functional NeuroImages)

​

NIMH

​

​

​

​

​

Statistical Parametric Mapping

​

The Wellcome Trust Centre for Neuroimaging, UCL

Link

https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FSL

https://afni.nimh.nih.gov/

http://www.fil.ion.ucl.ac.uk/spm/

Imaging

File Format

NIFTI

(and a little GIFTI)

NIFTI

(and GIFTI - SUMA)

NIFTI

Platform

GUI and linux (command-line) tools

GUI + linux (command-line) tools

Matlab Toolkit

13 of 23

fMRI data preprocessing -

stuff every package does..

​

removal of “dummy” TRs

Slice timing

Motion correction

Temporal filtering

Spatial Smoothing

Registration to MNI template

14 of 23

fMRI data preprocessing - removing noise...

  • Regressing out Physiological Signals
  • CompCor (AFNI or MATLAB)
  • ANATICOR (AFNI)
  • ICA based noise cleaning
    • (FSL - MELODIC, FIX or AROMA..)

​

Signal

Components

Noise

Components

15 of 23

Resting State: Seed Based Analysis

Lafaille et al. IASR. 2013

16 of 23

Resting State: Seed Based Analysis

​

17 of 23

Resting State: Independent Components Analysis

​

Lafaille et al. IASR. 2013

18 of 23

PET data preprocessing -

stuff every package does..

​

realign

Motion correction

Coregister with MRI

Normalize to MNI template

Spatial Smoothing

Statistics

19 of 23

Viewing MR images

FSLview

MRIcron

MRIcroGL

20 of 23

Human Connectome Project’s

Connectome Workbench

Connectome workbench is a viewer (wb_view) and a set of tools (wb_command) for volumes and surfaces

​

194 sub-commands in wb_command

21 of 23

Nilearn - makes pretty pictures (in python)

​

22 of 23

Scripting to put it all together

Nipype is a tools for

Combining different packages (e.g., ANTS, SPM, FSL, FreeSurfer, Camino, MRtrix, MNE, AFNI, Slicer), into workflows in python

You can combine command line tools from multiple packages into workflows using bash scripting.

​

​

Epitome (written by J. Viviano of the Kimel Lab) can help you combine AFNI and FSL.

Consider writing workflows in python

23 of 23

Come to Office Hours!

Every Tuesday, 1-2pm

Room: CS 163

​

Complete the post-workshop survey!

bit.ly/scwgmay2017

​

​

​

​

Get on wifi!

SSID: StaffDevices

Password: shxz@yf1dcck8877

​

​

​

​

Scientific Computing Working Group

https://camh-scwg.github.io/compucool-may-2017/