Quantitative analysis of EEG signals: Time-frequency methods and ...
Quantitative analysis of EEG signals: Time-frequency methods and ...
Quantitative analysis of EEG signals: Time-frequency methods and ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
2. Registration <strong>of</strong> 120 sweeps covering 2s time-windows with application <strong>of</strong> unimodal<br />
stimulation using a rectangular light-step centered in the visual eld at a distance<br />
<strong>of</strong> 1.5 m.<br />
3. Registration <strong>of</strong> 120 sweeps covering 2s time-windows with application <strong>of</strong> multimodal<br />
stimulation. Stimuli <strong>of</strong> (1) <strong>and</strong> (2) were applied simultaneously.<br />
After visual inspection <strong>of</strong> the data, 30 sweeps free <strong>of</strong> artifacts were selected for each<br />
modality for future <strong>analysis</strong>. A Wavelet Transform was applied to each single sweep<br />
using a quadratic B-Spline function as mother wavelet. After a ve octave wavelet<br />
decomposition, the components <strong>of</strong> the gamma b<strong>and</strong> 31 ; 62Hz were analyzed. For each<br />
subject the gamma components <strong>of</strong> the 30 single sweeps were averaged. Results for each<br />
subject were averaged obtaining a \gr<strong>and</strong> average".<br />
Statistical <strong>analysis</strong><br />
The wavelet coecients processed were the ones obtained after averaging the responses<br />
<strong>of</strong> the 30 single trials for each electrode <strong>and</strong> subject. Then, wavelet coecients<br />
were rectied <strong>and</strong> the maximum coecient was computed in a time window <strong>of</strong> 250ms<br />
post-stimulation. A multiple factor ANOVA (MANOVA) with repeated measures test<br />
was applied in order to compare dierences between modalities <strong>and</strong> electrodes. Furthermore,<br />
maximum wavelet components pre- <strong>and</strong> post-stimulation were compared with<br />
t-tests in order to analyze statistical signicance <strong>of</strong> the amplitude enhancements for<br />
each electrode <strong>and</strong> modality.<br />
4.6.3 Results<br />
Gamma wavelet coecients <strong>of</strong> one typical subject are shown in gure 22. In central<br />
locations gamma amplitude enhancements are more pronounced upon bimodal stimulation.<br />
In this case, pre-stimulus <strong>EEG</strong> amplitude is less than 5 (in arbitrary units) <strong>and</strong><br />
post-stimulus enhancements reach values up to 12. Further increases are seen in right<br />
frontal <strong>and</strong> right occipital electrodes, in the latter case appearing with some delay.<br />
The gr<strong>and</strong> average <strong>of</strong> all the subjects is shown in gure 23. With auditory stimulation<br />
enhancements are visible in the central electrode <strong>and</strong> more poorly dened in the occipital<br />
ones. Visual stimulation shows enhancement onlyinthecentral electrode. On the other<br />
h<strong>and</strong>, bimodal stimulation evokes signicantly higher enhancements than the other two<br />
modalities (p < 0:01 see also g. 24) <strong>and</strong> they are distributed all along the surface<br />
<strong>EEG</strong>, being more pronounced in the central location. Furthermore, bimodal responses<br />
are clearly not a linear superposition <strong>of</strong> the auditory <strong>and</strong> visual ones, this fact being<br />
very clear in central electrodes <strong>and</strong> specially in the frontal ones, where enhancements<br />
occur only upon bimodal stimulation.<br />
62