Contents

• List of Figures
• 1. Introduction
  • 1.1 Background
  • 1.2 Purpose of the Work
  • 1.3 Structure of the Thesis
  
  
• 2. Brain Imaging
  • 2.1 The Human Brain
    • 2.1.1 Structural Anatomy
    • 2.1.2 Functional Anatomy
    • 2.1.3 Auditory Signal Processing
  • 2.2 Evolution of Imaging Techniques
  • 2.3 Functional Magnetic Resonance Imaging
    • 2.3.1 Measuring Hemodynamic Responses to Stimuli
    • 2.3.2 Standard Preprocessing of Images
    • 2.3.3 Standard Analysis of fMRI Sequences
  • 2.4 Individual and Group Studies
  
  
• 3. Independent Component Analysis
  • 3.1 Motivation
  • 3.2 Mixture Model
  • 3.3 Estimating Independence
    • 3.3.1 FastICA Algorithm
    • 3.3.2 Estimation Errors
  • 3.4 Application to fMRI Data
    • 3.4.1 Spatial ICA
  
  
• 4. Variability of Independent Components
  • 4.1 Motivation and Sources of Variability
  • 4.2 Exploiting Variability
    • 4.2.1 Randomizing the Initial Conditions
    • 4.2.2 Resampling the Data
  • 4.3 Analyzing Consistency
    • 4.3.1 Multiple Runs of ICA
    • 4.3.2 Clustering the Estimates
    • 4.3.3 Properties of the Groups
  • 4.4 Interpreting the Variability
  
  
• 5. Visualization
  • 5.1 Relevance
  • 5.2 Showing Brain Activation
    • 5.2.1 Spatial Information
    • 5.2.2 Temporal Information
    • 5.2.3 Group Information
  • 5.3 Complete User Interface
  • 5.4 Medical Standards
  
  
• 6. Experimental Setup
  • 6.1 Real fMRI Data
    • 6.1.1 Auditory Stimulation
    • 6.1.2 Volume Acquisition
    • 6.1.3 Volume Preparation
  • 6.2 Individual Experiments
  • 6.3 Group Experiment
  
  
• 7. Results
  • 7.1 Individual Results
    • 7.1.1 Overview
    • 7.1.2 Components Related to Stimulus
    • 7.1.3 Components Revealing Artifacts
    • 7.1.4 Components with Strong Variability
    • 7.1.5 Other Interesting Components
    • 7.1.6 Spatial Relations Revealed by Variability
  • 7.2 Group Results
  
  
• 8. Conclusions
  • 8.1 Discussion
  • 8.2 Medical Relevance
  • 8.3 Future Research
  
  
• A. Mathematical Concepts
  • A.1 Principal Component Analysis
    • A.1.1 Eigenvalue Decomposition
    • A.1.2 Principal Components
    • A.1.3 Whitening
    • A.1.4 Reducing Dimension
  • A.2 Estimating Independence
    • A.2.1 Mutual Information
    • A.2.2 Negentropy
    • A.2.3 Non-Gaussianity
  • A.3 Fast Fixed-Point Iteration
    • A.3.1 Deflation Mode
    • A.3.2 Symmetric Mode
  
  
• B. Complete Results
  • B.1 Description
  • B.2 Figures
  
  
• Bibliography