Understanding the brain as a network of interconnected nodes is the concept at the heart of connectomics. There have been many fascinating insights into the structure of this network and how it supports brain function in terms of the signals that researchers are able to measure, such as the BOLD signal in fMRI or electromagnetic activity in M/EEG. In the latter case, one can potentially gain insight into the time scale (milliseconds) relevant for behavior and neural events, especially when signals recorded on the scalp are successfully projected into the gray matter. However, the low spatial resolution and signal-to-noise ratio make it difficult to apply connectomics approaches to this kind of data. In our recent work, we integrate white matter connectivity data with high-density EEG recordings using a well-suited analysis framework called graph signal processing (GSP). GSP allows us to extract basis functions of the white matter connectivity - called "network harmonics" - and use them as building blocks for EEG activity patterns. In my talk, I will introduce this method including some of its links to harmonic modes in other areas of science. I will show how it yields a sparse representation of the EEG signal which allows us to track fast spatio-temporal dynamics over the course of a simple visual task.
Send your questions to our Q & A Space: https://neurostars.org/t/workshop-spatiotemporal-dynamics-in-neuroimaging-models-and-analysis-q-a/7608?u=psanzleon
