Cross-frequency Granger causality

Dear Mike,
I’ve been working on Granger Causality analysis following your book and videos. I plan to test whether posterior lateral alpha power can be used to predict frontal theta power. I know using the filtered lateral alpha and frontal theta power time series to calculate Granger causality is not recommended. However, I also noticed that spectral Granger causality was usually computed at the same frequency for both brain regions instead of across different frequencies. In my case, lateral alpha power can only be obtained by calculating the difference wave between contralateral and ipsilateral electrodes. Common spectral Granger causality analysis seemed to be not applicable herein. May I ask for your suggestions on my data? Or there’s other way to test my hypothesis?

Thanks a lot!

Hi Wen. To my knowledge, there is no cross-frequency GC measure. However, there are methods to measure directionality between the low- and high-frequency activity, and you can combine that with using different electrodes for the phase and power. Together, I believe that will do what you’re looking for.

Here’s the relevant paper: