Bioimpedance Spectroscopy Methods for Analysis and Control of Neurostimulation Dose

Abstract

TDCS is a form of non-invasive neurostimulation that is comprised of injection of current via strategically placed scalp electrodes into targeted areas of the brain. TDCS has shown therapeutic benefit for numerous clinical applications. This technique has not however been widely adopted due to high variability of response to the stimulation. Current state of the art methods for optimizing tDCS are based on FEM models that generally model tissue as isotropic and homogeneous and do not take into account inter subject variability of head tissue electrical properties. We therefore develop an in-vivo method of measuring and analyzing bioimpedance spectroscopy measurements of the head to estimate change to tDCS dose in neural tissues for different subjects.

Finite element simulations are implemented on a realistic MRI derived head model. 5\% random Gaussian noise is added. Experimental bioimpedance measurements are taken of the heads of 8 subjects. We simulate sensitivity distribution and impedance for a variety of 2 and 4 electrode configurations over a wide frequency range. We also extract Cole parameters and implement PCA on simulated and experimental impedance.

We demonstrate that the Cole model of the head can be accurately approximated by the sum in series of Cole systems of each tissue. Comparison of Cole parameters from various simulated electrode configurations show statistical differences (paired t test $p<.05$). PCA shows that close to 100\% of the variance between two impedance spectra is described along a single principal component. Variation described by the second principal component increases as a function of increasing inter electrode gap which may be related to changes in dose. FEM and experimentally derived Cole parameters show different trends for various electrode configurations, good agreement is however shown for the PCA results. The outcome of this research may lead to a higher tDCS efficacy by improving standardization and control of stimulation by relation of dose and bioimpedance spectra characteristics.