A hybrid analytical–numerical algorithm for determining the neuronal current via electroencephalography

Abstract

In this study, the neuronal current in the brain is represented using Helmholtz decomposition. It was shown in earlier work that data obtained via electroencephalography (EEG) are affected only by the irrotational component of the current. The irrotational component is denoted by Ψ and has support in the cerebrum. This inverse problem is severely ill-posed and requires that additional constraints are imposed. Here, we impose the requirement of the minimization of the L 2 norm of the current (energy). The function Ψ is expanded in terms of inverse multiquadric radial basis functions (RBF) on a uniform Cartesian grid inside the cerebrum. The minimal energy constraint in conjunction with the RBF parametrization of Ψ results in a Tikhonov regularized solution of Ψ. The RBF shape parameter (regularization parameter), is computed by solving a 1-D non-linear maximization problem. Reconstructions are presented using synthetic data with a signal to noise ratio (SNR) of 20 dB. The root mean square error (RMSE) between the exact and the reconstructed Ψ is RMSE=0.1122. The proposed reconstruction algorithm is computationally efficient and can be vectorized in MATLAB.

Time Perception through the Processing of Verb Tenses: An ERP study regarding Mental Time Travel

“…This algorithm uses real brain topology without the spherical approximation. Due to certain technical difficultie s, it was not possible to use the latter algorithm in the present work; however, there is a broad agreement between the results obtained via the LORETA technique and the approach of this algorithm (Hashemzadeh et al, 2020).…”

“…Based on EEG data, it is not possible to reconstruct the neuronal current uniquely (Dassios and Fokas, 2020). A novel algorithm to reconstruct the 'visible' by EEG part of the current, namely the part of the current that affects the EEG data, is presented in a recent study (Hashemzadeh et al, 2020). This algorithm uses real brain topology without the spherical approximation.…”

Vector-valued spline method for the spherical multiple-shell electro-magnetoencephalography problem

“…However, the method allows new insights to the non-uniqueness of the underlying problem, for instance, a direction of the neuronal current is invisible for either the MEG as well as the EEG measurements. A first attempt of transferring results for the analytical model to more realistic brain geometry is done in [33].…”

“…In the context of approximations of the neuronal current based on continuously distributed MEG and EEG models, a regularization method based on global orthonormal basis functions (related to spherical harmonics) and scalar spherical splines have been used before [22]. In addition, for the EEG model a hybrid analytical-numerical algorithm using OpenMEG exists [33]. Besides, discrete models (e.g.…”

Can artificial intelligence reach human thought?

“…The numerical implementation of this algorithm requires the computation of a certain auxiliary function ( 9 , 10 ). This well-defined goal can be achieved via the training of a two-layer neural network, which provides yet another illustration of the importance of machine learning ( 11 ). The solution of this specific goal is conceptually very different to my work on the Lindelöf hypothesis ( 12 ), where a novel approach was introduced to this historical problem.…”