Loading...

Quantitative Analysis of Epileptic Seizure EEG

Hoseini, Mahmood | 2010

461 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 41130 (04)
  4. University: Sharif University of Technology
  5. Department: Physics
  6. Advisor(s): Rahimi Tabar, Mohammad Reza
  7. Abstract:
  8. Since recording first electroencephalogram (EEG) of human brain in 1929 until now it becomes as a powerful tool in neuroscience. At first information extraction was done by visionary approaches only. But because of some problems in the context of inaccuracy and also in analyzing data different methods were proposed in order to extract hidden information of EEG. Among these approaches Fourier transformation was suggested as a very useful method that could draw out so many characteristics of signal different frequency components. However this way had many faults that cause limitation in analyzing time series and as a result other methods have been considered. One method that later has been considered in fact was time-frequency form of Fourier transformation that nominated as Gabor transformation. In this thesis we are going to discuss this latter approach and will express its advantages in comparison to Fourier transformation. Then by relating its deficiencies, we will come across a more applicable time-frequency way by the name Wavelet transformation. Wavelet transformation gives us very much useful information about dynamics and focus of epilepsy whereas we don’t have access to that information by using Fourier analysis. At the end we analyze EEG signals recorded from epileptic patients by this approach and also try to determine phase velocity of neural pulses in brain by computing correlation function of different channels in different scales. We also make our efforts in determining the focus of epilepsy and synchronization of electrodes in ictal period
  9. Keywords:
  10. Epilepsy ; Electroencphalogram ; Time Series ; Wavelet Transform

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC