Inter-trial phase clustering

[@thedarknightnicht]

What is the solution for changing the direction of the distribution in the Matlab part?

[@kash983]

Preferred Angle represents what? What is the physical interpretation of the 'prefAngle'?

[@adamattaheri8434]

Thank you Mike, I am finding your lecture series very useful. I do have one question if you would be so kind. Taken together with the information in another of your lectures kzbin.info/www/bejne/b3nOXoV3ft-Cl9k&ab_channel=MikeXCohen I am getting confused about how you have ITPC values at the edges of your data analysis window. Using the wavelet I believe we are computing the dot product between the Kernal and a portion of the EEG signal (which gives one result for a particular wavelet kernel and a portion of EEG data). I also believe that you should align the center of your wavelet kernel to the time point of interest (e.g. the EEG aligned to a stimulus onset). However, in the above Matlab example, your EEG data goes from minus 1000ms to 1500ms. I believe zero-padding gets around this problem by creating "wings of convolution". Therefore are the phase values at -1000ms the result of a dot product of the right-hand half of the wavelet aligned to real data and the left-hand half of the wavelet aligned to zeros (added via zero padding)? Furthermore, is this data less reliable than the including pre-trial onset EEG data (if available) instead of zero padding? Many thanks for any clarification.

[@mikexcohen1]

Hi Adam. Good question. There are always edge effects from filtering, any data close to the edges should never be overly trusted. In practice, you would ignore any results close to the edges of the signal. That goes for time-domain filtering, power, ITPC, and any other result of filtering. This is what I call "buffer zones" in other videos (I think it's the one about preprocessing).

[@DemetrisRoumis]

13:15 equation labels misaligned