During rest different brain structures are connected into distinguishable networks. Each region of interest (ROI) tends to preserve its own natural frequency. A long term goal of our team is to develop a reliable method which al‑ lows us to access milliseconds-level dynamic features of brain networks from EEG signal, so called EEG brain finger‑ print. In 2016, Dr. Anne Keitel and Professor Joachim Gross proposed a method of clustering each ROI MEG source-re‑ constructed activity. 116 ROIs were selected according to the Automated Anatomic Labeling Atlas. As a result, they obtained a group-level spectral representation (referred to as “spectral fingerprints”) of dynamic behavior of the hu‑ man cortex and deep sources. It turned out that spectral fingerprint representation enables accurate ROI classifica‑ tions. Moreover, using only functional data, the algorithm linked ROIs to clusters that correspond well to large-scale anatomical parcellations of the cortex. We have recreat‑ ed and adapted this method to analyse resting-state EEG signal. We tested this approach on resting-state EEG data and compared with original results obtained on MEG data. Results show that, despite the difficulty of differentiating between ROIs using EEG data, performance was well above the chance level. Each ROI EEG activity was characterized by fewer numbers of clusters than ROI MEG activity and ROI similarities formed less clear images of brain networks. Nevertheless, the method is still promising and further de‑ velopment could lead us to identify new and reliable tools useful in scientific and clinical practice.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.