Remotely induced electrical modulation of deep brain circuits in non-human primates

Abstract

Introduction: The combination of magnetic and focused ultrasonic fields generates focused electric fields at depth entirely noninvasively. This noninvasive method may find particularly important applications in targeted treatments of the deep brain circuits involved in mental and neurological disorders. Due to the novelty of this method, it is nonetheless unknown which parameters modulate neural activity effectively.

Methods: We have investigated this issue by applying the combination of magnetic and focused ultrasonic fields to deep brain visual circuits in two non-human primates, quantifying the electroencephalographic gamma activity evoked in the visual cortex. We hypothesized that the pulse repetition frequency of the ultrasonic stimulation should be a key factor in modulating the responses, predicting that lower frequencies should elicit inhibitory effects and higher frequencies excitatory effects.

Results: We replicated the results of a previous study, finding an inhibition of the evoked gamma responses by a strong magnetic field. This inhibition was only observed for the lowest frequency tested (5 Hz), and not for the higher frequencies (10 kHz and 50 kHz). These neuromodulatory effects were transient and no safety issues were noted.

Discussion: We conclude that this new method can be used to transiently inhibit evoked neural activity in deep brain regions of primates, and that delivering the ultrasonic pulses at low pulse repetition frequencies maximizes the effect.

Keywords: Lorentz force; incisionless; induction; magnetic field; neuromodulation; noninvasive; ultrasound.

Figure 1

Concept and experimental setup. (A) Remus stimulation system and Lstim concept. An ultrasonic transducer array was attached to titanium pins mounted in the animal’s skull. This fixed positioning enabled us to deliver ultrasound into a deep brain target, the lateral geniculate nucleus (LGN) repeatably from session to session. An ultrasound wave, focused into a target with acoustic impedance Z, induces in the target motions of molecules with velocity V=$\frac{P}{Z}$. The pressure P (and so the velocity V) are maximal at the target. When the wave is emitted in a direction perpendicular to the magnetic field B, so that the velocity vector is perpendicular to B, the target experiences a localized electric field E=$\frac{\mathit{PB}}{Z}$. (B) Manipulation of the magnetic field. The animal laid on an MRI table in standard sphinx position. The table allowed us to expose the animal’s brain to either high field (3T) at the isocenter of the bore and nearly zero field (0T) 3m away from the isocenter. (C) Validation of the ultrasound targeting. MRI thermometry was performed in each animal to validate targeting of the LGN. The details of this validation with these animals are published (Webb et al., 2022; Webb T. D. et al., 2023).

Figure 2

Modulation of the deep brain target by ultrasound and Lstim. Mean ± s.e.m. gamma power change relative to the 1 s baseline before stimulus onset. Shaded error bars represent the standard error of the mean across sessions at each individual time point. The vertical gray bars represent the time of the ultrasonic stimulation (1,000 ms for 5 Hz PRF and 100 ms for 10 and 50 kHz PRF). The horizontal black bars represent the analysis window. *: p < 0.05, two-sided Wilcoxon rank-sum test with a Bonferroni correction.

Figure 3

Summary of the results. Median ± 25-75th percentiles of the data (Figure 2) in the analysis window. We used these nonparametric representations and tests as the data were not necessarily normally distributed.*: p < 0.05; two-sided Wilcoxon rank-sum test with Bonferroni correction.

Figure 4

Modulation of the deep brain target by ultrasound and Lstim, separated by animal. Mean ± s.e.m. gamma power change relative to the 1 s baseline before stimulus onset. Shaded error bars represent the standard error of the mean across sessions at each individual time point. The vertical gray bars represent the time of the ultrasonic stimulation (1,000 ms for 5 Hz PRF and 100 ms for 10 and 50 kHz PRF). The horizontal black bars represent the analysis window. *: p < 0.05, two-sided Wilcoxon rank-sum test with a Bonferroni correction.