Structure-function coupling of the video-watching EEG on the underlying anatomy

Poster Session F - Tuesday, April 16, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Venkatesh Subramani^1,2 (venkatesh.subramani@umontreal.ca), Giulia Lioi², Karim Jerbi¹, Nicolas Farrugia²; ¹Université de Montréal, Canada, ²IMT Atlantique, France

Brain Structure-Function relationship (i.e. between activity and structural connectivity) is well characterized for fMRI[1,2], and has been studied with EEG during epileptic seizures and event-related potentials[3,4], but has been seldom investigated using continuous EEG. Here, we investigate the structure-function coupling during movie-watching using EEG and MRI data from Healthy Brain Network (n=43)[5] with the resting-state as baseline reference and exploiting the graph signal processing(GSP) framework[6]. EEG cortical sources were estimated using Boundary Element Method and eLORETA, followed by parcellation on the HCP-MMP atlas, Hilbert transform and bandpass filtering in specific frequency bands. Cortical envelopes were projected to the subject-specific structural connectome[7] and the structure-function relationship was quantified using a GSP metric (Structural-decoupling Index: SDI)[8]. During movie-watching, RetroInsular Cortex, Visual V3, and Primary Auditory cortex are closely aligned with anatomy, while the Parahippocampal and inferior frontal sulcus areas are decoupled. The reliability of this pattern was ensured with a different video (Intraclass coefficients: 0.78; 95%CI: 0.72, 0.83). Comparing video and rest SDI, we observed weaker coupling during movie-watching in many regions such as the Premotor and Fusiform Face complex, while the Retrosplenial complex in the PCC and Premotor cortex exhibited stronger coupling. In line with previous work on fMRI[9,8], EEG in sensory areas is more strongly coupled to the underlying structure, while transmodal regions such as PCC exhibit strong coupling. Overall, this study provides novel insights on how the dynamics of continuous EEG relate to the individual brain structure during a complex, naturalistic task as compared to resting state.

Topic Area: METHODS: Electrophysiology