An EEG Finger-Print of fMRI deep regional activation

​

/ Meir-Hasson Y, Kinreich S, Podlipsky I, Hendler T, Intrator N.
Neuroimage. 2014 Nov 15;102 Pt 1:128-141.

 

Introducing the EEG finger-print (EFP); a computational model that tracks the activity of deep brain regions via EEG, thereby reducing the need for fMRI scanning. 

Traditional EEG neurofeedback paradigms, such as alpha/theta sampling, provide low spatial accuracy, which makes deep brain regions, such as the amygdala, largely inaccessible. The EFP model was developed in order to overcome this limitation, by using simultaneous EEG-fMRI recordings. It exploits the high spatial resolution imaging enabled by fMRI, and incorporates advanced signal processing and machine learning methods of EEG signals. Thereby, the model learns to predict a brain region’s activity as measured with the fMRI, based on the signals recorded in EEG.

 

An amygdala-EFP model successfully predicted the amygdala’s fMRI signal from a single EEG electrode. Moreover, it provided better prediction of amygdala activity than the traditional alpha/theta EEG sampling. Thus, the EFP is proposed as a more targeted biomarker of neural activity, which can be applied in EEG-based neurofeedback and other brain-guided procedures. 

​

To read the full article in NeuroImage -https://www.sciencedirect.com/science/article/abs/pii/S1053811913010963?via%3Dihub