Loading...

Blind Source Separation Analysis of brain fMRI for Activation Detection

Akhbari, Mahsa | 2010

646 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 40452 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Fatemizadeh, Emadeddin; Babaiezadeh, Massoud
  7. Abstract:
  8. Functional Magnetic Resonance Imaging (fMRI) is one of the imaging techniques that are used to study human brain function and neurological disease diagnosis. Popular techniques in fMRI utilize the blood oxygenation level dependent (BOLD) contrast, which is based on the differing magnetic properties of oxygenated (diamagnetic) and deoxygenated (paramagnetic) blood. In order to analyze fMRI data, hypothesis-driven or data-driven methods can be used. Among data-driven techniques, Independent Component Analysis (ICA) provides a powerful method for the exploratory analysis of fMRI data. In this thesis, we use ICA on fMRI data for detecting active regions in brain, without a-priori knowledge of neural stimulus. We propose two criterions based on smoothness and based on the estimation of the entropies of time courses corresponding to component maps, to select automatically the components of interest. And also in this thesis, we apply “window ICA” in order to find meaningful components and active regions in different time intervals. We also modify the estimated mixing matrix in an ICA algorithm in order to solve the indeterminacy of ICA and find components in a predefined order. At last we use a method based on Sparse Component Analysis (SCA) method to find active regions of fMRI data
  9. Keywords:
  10. Blind Sources Separation (BSS) ; Independent Component Analysis (ICA) ; Functional Magnetic Resonance Imaging (FMRI) ; Activation Detection ; Sparse Component Analysis (SCA) ; Entropy ; Component Time Order

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC