Module 8

Module 8: Pre-processing I
Data Processing Pipeline

! !Experimental Design !
Data Acquisition !
Preprocessing ! !
Reconstruction!
Slice-time ! ! ! Correction! ! ! !Motion Correction,! ! Co-registration &! ! !Normalization! ! ! ! Spatial! ! Smoothing! !
Data Analysis! ! ! Localizing !

! Brain Activity! ! ! ! ! Connectivity! ! ! ! ! ! Prediction! ! !
Pre-processing
! Prior to analysis, fMRI data undergoes a series of preprocessing steps aimed at identifying and removing artifacts and validating model assumptions. ! The goals of preprocessing are
! To minimize the influence of data acquisition and physiological artifacts; ! To check statistical assumptions and transform the data to meet assumptions; ! To standardize the locations of brain regions across subjects to achieve validity and sensitivity in group analysis.
Pre-processing Pipeline
Structural (T1)
Coregister to func Warp to atlas template Warping parameters
Apply
T1 in atlas space
Functional image time series

! Denoise ! Slice timing (a) ! Realign (r) Smooth
Preprocessing is performed both on the fMRI data and structural scans collected prior to the experiment.
Pre-processing Steps
! Visualization and Artifact Removal ! Slice Time Correction ! Motion Correction ! Physiological Corrections ! Co-registration ! Normalization ! Spatial Filtering ! Temporal Filtering
Visualization & Artifact Removal

! The first part of the preprocessing pipeline is to use exploratory techniques to investigate the raw image data and detect possible problems and artifacts. ! fMRI data often contain transient spike artifacts and slow drift over time. ! An exploratory technique such as principal components analysis (PCA) can be used to look for spike-related artifacts.
Transient spike artifacts in the data during isolated volume acquisitions are apparent in certain slices, as shown by the bright bands in the saggital slices (bottom). This suggests that gradient performance was affected during acquisition of some echo-planar images, which were acquired slice-byslice in interleaved order in this experiment.
Slice Time Correction

! We often sample multiple slices of the brain during each individual repetition time (TR) to construct a brain volume. ! Typically each slice is sampled at a slightly different time points (i.e., 2D imaging; not 3D). ! Slice time correction shifts each voxel's time series so that they all appear to have been sampled simultaneously.

Slice 1 Slice 2 Slice 3 1 2 3 4 5 ..

Slice 1 Slice 2 Slice 3 1 2 3 4 5 ..
Can be corrected using temporal interpolation.

1 2 3 4 5
Temporal Interpolation ! Use information from nearby time points to estimate the amplitude of the MR signal at the onset of the TR." Use a linear, spline or sinc function."
Phase Shift ! Slide the time course by applying a phase shift to the Fourier transform of the time course.
Head Motion
! Very small movements of the head during an experiment can be a major source of error if not treated correctly. ! When analyzing the time series associated with a voxel, we assume that it depicts the same region of the brain at every time point
! Head motion may make this assumption incorrect.
! Can be corrected using a rigid body transformation.
Motion Correction
! The goal is to find the best possible alignment between an input image and some target image. ! To align the two images, one of them needs to be transformed. ! A rigid body transformation is used. ! It involves 6 variable parameters, 3 sets of translations and 3 sets of rotations (6 DOF).
Translation
Rotation
Scaling
Shearing
Transformations
Linear transformations ! Rigid body (6 DOF) translation and rotation ! Similarity (7 DOF) translation, rotation and a single global scaling ! Affine (12 DOF) translation, rotation, scaling and shearing. Warping Transformations where the equations relating the coordinates of the images are non-linear.
Motion Correction
! The target image is usually defined to be the first (or mean) image in the fMRI time series. ! The goal is to find the set of parameters which minimizes some cost function that assesses similarity between the image and the target. ! Examples of cost functions include the sum of squared differences or mutual information.
Illustration
Coregistration
! Next, a structural MRI collected in the beginning of the session is registered to the fMRI images in a process referred to as coregistration.
! Allows one to visualize single-subject task activations overlaid on the individuals anatomical information. ! Simplifies later transformation of the fMRI images to a standard coordinate system.
Coregistration
! There are certain key differences between coregistration and motion correction.
! Functional and structural images do not have the same signal intensity in the same areas. !! They cannot be subtracted. ! Their shapes may differ.
! Use at least an affine transformation to perform coregistration and the mutual information cost function.
Coregistration
End of Module
@fMRIstats

Module 8

Uploaded by

Copyright:

Available Formats

Module 8

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Module 8

Uploaded by

Copyright:

Available Formats

Module 8: Pre-processing I

Data Processing Pipeline

Slice-time ! ! ! Correction! ! ! !Motion Correction,! ! Co-registration &! ! !Normalization! ! ! ! Spatial! ! Smoothing! !

Data Analysis! ! ! Localizing !

Functional image time series

Visualization & Artifact Removal

Slice Time Correction

Slice Time Correction

Slice Time Correction

Can be corrected using temporal interpolation.

! Can be corrected using a rigid body transformation.

You might also like