About the book
Table of Contents
1 Introduction
1.1 A brief overview of fMRI
1.1.1 Blood flow and neuronal activity
1.1.2 Magnetic resonance imaging
1.2 The emergence of cognitive neuroscience
1.3 A brief history of fMRI analysis
1.4 Major components of fMRI analysis
1.5 Software packages for fMRI analysis
1.5.1 SPM
1.5.2 FSL
1.5.3 AFNI
1.5.4 Other important software packages
1.6 Choosing a software package
1.7 Overview of processing streams
1.8 Prerequisites for fMRI analysis
2 Image Processing
2.1 What is an image?
2.2 Coordinate systems
2.2.1 Radiological and neurological conventions
2.2.2 Standard coordinate spaces
2.3 Spatial transformations
2.3.1 Transformation models
2.3.2 Cost functions
2.3.3 Estimating the transformation
2.3.4 Reslicing and interpolation
2.4 Filtering and Fourier analysis
2.4.1 Fourier Analysis
2.4.2 Filtering
2.4.3 Convolution
3 Preprocessing
3.1 Introduction
3.2 An overview of fMRI preprocessing
3.3 Quality control techniques
3.3.1 Detecting scanner artifacts
3.3.2 Timeseries animation
3.3.3 Independent components analysis (ICA)
3.4 Distortion correction
3.5 Slice timing correction
3.6 Motion correction
3.6.1 Motion correction techniques
3.6.2 Prospective motion correction
3.6.3 Quality control for motion correction
3.6.4 Interactions between motion and susceptibility artifacts
3.6.5 Interactions between motion correction and slice timing correction
3.6.6 How much is too much motion?
3.6.7 Physiological motion
3.7 Spatial smoothing
3.7.1 How much should I smooth?
4 Normalization
4.1 Introduction
4.2 Anatomical variability
4.3 Coordinate spaces for neuroimaging
4.4 Atlases and templates
4.4.1 The Talairach Atlas
4.4.2 The MNI templates
4.5 Preprocessing of anatomical images
4.5.1 Bias field correction
4.5.2 Brain extraction
4.5.3 Tissue segmentation
4.6 Processing streams for fMRI normalization
4.7 Spatial normalization methods
4.7.1 Landmark-based methods
4.7.2 Volume-based registration
4.7.3 Computational anatomy
4.8 Surface-based methods
4.9 Choosing a spatial normalization method
4.10 Quality control for spatial normalization
4.11 Troubleshooting normalization problems
4.12 Normalizing data from special populations
4.12.1 Normalizing data from children
4.12.2 Normalizing data from the elderly
4.12.3 Normalizing data with lesions
5 Statistical Modeling
5.1 The BOLD signal
5.1.1 Convolution
5.1.2 Beyond the canonical HRF
5.1.3 Other Modeling Considerations
5.2 The BOLD noise
5.2.1 Characterizing the noise
5.2.2 High-pass filtering
5.2.3 Prewhitening
5.2.4 Precoloring
5.3 Study design and Modeling Strategies
5.3.1 Study design: Estimation and detection
5.3.2 Study design: Multiple stimulus types
5.3.3 Optimizing fMRI designs
6 Statistical Modeling: Group analysis
6.1 The mixed effects model
6.1.1 Motivation
6.1.2 Mixed effects modeling approach used in fMRI
6.1.3 Fixed effects models
6.2 Mean centering continuous covariates
6.2.1 Single group
6.2.2 Multiple groups
7 Statistical Inference
7.1 Basics of statistical inference
7.2 Features of Interest in Images
7.2.1 Voxel-level inference
7.2.2 Cluster-level inference
7.2.3 Set-level inference
7.3 The Multiple Testing Problem & Solutions
7.3.1 Familywise Error Rate (FWE)
7.3.2 False Discovery Rate (FDR)
7.3.3 Inference Example
7.4 Combining Inferences: Masking & Conjunctions
7.5 Use of Region of Interest Masks
7.6 Computing Statistical Power
8 Connectivity
8.1 Introduction
8.2 Functional connectivity
8.2.1 Seed voxel correlation: Between-subjects
8.2.2 Seed voxel correlation: Within-subjects
8.2.3 Beta-series correlation
8.2.4 Psychophysiological interaction
8.2.5 Multivariate decomposition
8.2.6 Partial least squares
8.3 Effective connectivity
8.3.1 Challenges to causal inference with fMRI data
8.3.2 Path analysis and structural equation modeling
8.3.3 Graphical causal models
8.3.4 Dynamic causal modeling
8.3.5 Granger causality
8.4 Network analysis and graph theory
8.4.1 Small-world networks
8.4.2 Modeling networks with resting-state fMRI data
8.4.3 Preprocessing for connectivity analysis
9 Pattern Analysis
9.1 Introduction to pattern classification
9.1.1 An overview of the machine learning approach
9.2 Applying classifiers to fMRI data
9.3 Data extraction
9.4 Feature selection
9.5 Training and testing the classifier
9.5.1 Feature selection/elimination
9.5.2 Classifiers for fMRI data
9.5.3 Which classifier is best?
9.5.4 Assessing classifier accuracy
9.6 Characterizing the classifier
10 Visualization
10.1 Visualizing activation data
10.2 Localizing activation
10.2.1 The Talairach atlas
10.2.2 Anatomical atlases
10.2.3 Probabilistic atlases
10.2.4 Automated anatomical labeling
10.3 Localizing and reporting activation
10.3.1 Reporting Brodmann's areas.
10.3.2 Creating coordinate lists
10.4 Region of interest analysis
10.4.1 ROIs for statistical control
10.4.2 Defining ROIs
10.4.3 Quantifying signals within an ROI
Appendices
A GLM Intro/Review
A.1 Estimating GLM parameters
A.1.1 Simple linear regression
A.1.2 Multiple linear regression
A.2 Hypothesis Testing
A.3 Correlation and heterogeneous variances
A.4 Why "general" linear model?
B Data Organization and Management
B.1 Computing for fMRI analysis
B.2 Data organization
B.3 Project management
B.4 Scripting for data analysis
B.4.1 Some nuggets for scripting fMRI analyses
C Image formats
C.1 Data Storage
C.2 File Formats
C.2.1 DICOM
C.2.2 Analyze
C.2.3 NIfTI