NetStruct

INFERRING CEREBRAL NETWORK STRUCTURE: A STRATEGY TO IDENTIFY NETWORK INTERACTION POINTS FOR THE DECODING AND MODULATION OF BRAIN ACTIVITY


Relevant for Research Area

A - Foundations

 

 


Summary

Tools to record brain activity at high resolution and to focally stimulate the brain based on the recorded signals are the basis of novel strategies of therapeutical intervention in patients. Individual brain areas do not function in isolation; they rather form a tight mesh of many dynamically interacting cortical and subcortical regions. The main objective of our project is to identify these networks from recorded signals and suggest efficient ways to interfere with them in a bidirectional fashion. Of crucial importance for the safe application of our new reconstruction method to human brain networks is a trustworthy validation of its performance. We are testing the procedure for two recording methods that optimally complement each other: Electrocorticography (ECoG) and MR-Encephalography (MREG). Combining the two methods will enable a direct integration of networks at different scales: global (brain-wide) network structures with MREG, and more local and fine-scale ones with ECoG.


Research Status

We developed a novel method to infer the underlying effective network structure from measurements of human brain activity and tested it for ECoG and MREG applications. In the case of ECoG, we used simulations of biologically realistic spiking networks to verify the reconstruction performance. In the case of MREG, we employed simulations of autoregressive processes for calibration, and compared the outcome of diffusion tensor imaging (DTI) with the networks inferred from dynamic MREG signals.

Project Publications

Kern M et al (2013) Heart cycle-related effects on event-related potentials, spectral power changes, and connectivity patterns in the human ECoG. NeuroImage 81: 178–190

Lennartz C, Schiefer J, Rotter S, Hennig J, LeVan P (2018) Sparse Estimation of Resting-State Effective Connectivity from fMRI Cross-Spectra. Frontiers in Neuroscience 12: 287

Pernice V, Rotter S (2013) Reconstruction of sparse connectivity in neural networks from spike train covariances. Journal of Statistical Mechanics P03008

Proulx S, Safi-Harb M, LeVan P, An D, Watanabe S, Gotman J (2014). Increased sensitivity of fast BOLD fMRI with a subject-specific hemodynamic response function and application to epilepsy. NeuroImage 93: 59-73

Schiefer J, Niederbühl A, Pernice V, Lennartz C, LeVan P, Hennig J, Rotter S (2018) From Correlation to Causation: Estimation of Effective Connectivity from Continuous Brain Signals based on Zero-Lag Covariance. PLOS Computational Biology 14(3): e1006056