In-vitro Recordings of Neural Magnetic Activity From the Auditory Brainstem Using Color Centers in Diamond - A Simulation Study

At a Glance

Metadata	Details
Publication Date	2021-05-13
Journal	Frontiers in Neuroscience
Authors	Mürsel Karadas, Christoffer Olsson, Nikolaj Winther Hansen, Jean‐Françóis Perrier, James L. Webb
Institutions	Hvidovre Hospital, Copenhagen University Hospital
Citations	5
Analysis	Full AI Review Included

In-vitro Recordings of Neural Magnetic Activity Using Color Centers in Diamond: A Simulation Study

Executive Summary

Core Value Proposition: This simulation study validates the feasibility of using Nitrogen-Vacancy (NV) center magnetometry in diamond for stain-free, high-spatial-resolution 2D imaging of fast action potentials (APs) in neural tissue (mouse auditory brainstem slice).
Feasibility Conclusion: High-resolution magnetic imaging of individual APs requires next-generation NV sensors, necessitating a sensitivity improvement of two orders of magnitude (target: 10 nT µm).
Current Capability: Existing NV sensors possess sufficient sensitivity (1,520 nT µm) to support the magnetic sensing of cumulated neural signals sampled from larger areas (mm-scale spatial averaging).
Stimulation Comparison: Electrical stimulation generated a peak magnetic field of ~0.5 nT, approximately five times stronger than the ~0.1 nT peak field generated by optogenetic stimulation.
Optogenetic Advantage: While weaker, optical stimulation is favorable for avoiding electrical artifacts in magnetic recordings and is confirmed to be physiologically safe, with tissue heating remaining below 2°C even under high-power pulsed conditions (4 W/mm²).
Pathway Dominance: The magnetic signal is dominated by the axial currents flowing along the myelinated axons of the Globular Bushy Cells (GBCs) rather than the signals originating from the Calyces of Held or the MNTB principal cell somas.

Technical Specifications

Parameter	Value	Unit	Context
Target High-Resolution Sensitivity	10	nT µm	Required for spatially resolved AP imaging
Existing NV Sensor Sensitivity (Area-Normalized)	1,520	nT µm	Based on 5 µm NV layer, 10 kHz sampling rate
Existing NV Sensor Sensitivity (Volume-Normalized)	34	nT mHz^-1/2	Continuous wave protocol
Peak Magnetic Field (Electrical Stim)	~0.5	nT	Worst-case estimate (300 cells, 25 µm distance)
Peak Magnetic Field (Optical Stim)	~0.1	nT	Worst-case estimate (300 cells, 25 µm distance)
Brain Slice Thickness (Active Layer)	300	µm	Simulated mouse brainstem slice
NV Sensor Distance to Active Cells	25	µm	Accounts for inactive cell layer due to cutting
Max Temperature Rise (Pulsed Stim)	<2	°C	4 W/mm² source irradiance, 40 Hz frequency
Light Wavelength (ChR2 Excitation)	473	nm	Optogenetic stimulation wavelength
Light Irradiance (Source, High Power)	4	W/mm²	Extreme case optical stimulation
Brain Tissue Thermal Conductivity (k)	0.56	W/mK	Used in Pennes bioheat equation
Brain Tissue Specific Heat (c)	3.6	J/gK	Used in Pennes bioheat equation
ChR2 Channel Density (Assumed)	1.3e10	channels/cm²	Based on bacteriorhodopsin expression in oocyte

Key Methodologies

The study utilized a multi-physics simulation approach combining neural dynamics, optogenetics, light transport, and magnetic field modeling:

Neural Dynamics Modeling (NEURON):
- The GBC-MNTB pathway was modeled using the cable equation to calculate membrane potentials (V_m) and transmembrane currents (I_m) with a time resolution of 25 µs.
- Cell models incorporated Hodgkin-Huxley-like dynamics, including inactivating Na⁺, low-threshold K⁺ (K_LT), and high-threshold K⁺ (K_HT) channels.
- Synaptic transmission at the Calyx of Held was modeled using two kinetics: vesicle release (500 active zones) and a 6-state AMPA receptor gating model.
Optogenetic Channel Modeling (ChR2):
- Channelrhodopsin-2 (ChR2) was implemented using a four-state kinetic model (two closed, two open states) to simulate photo-current generation.
- Light-dependent transition rates (K_a1, K_a2) were dynamically adjusted based on the photon flux (irradiance) during illumination.
Light Distribution and Transport Modeling:
- The Kubelka-Munk (KM) model was primarily used to estimate light irradiance (I) within the brain tissue slice, accounting for geometric spread, scattering, and absorption.
- Monte Carlo (MC) simulations were performed to obtain robust, worst-case estimates for light absorption used in thermal analysis.
Thermal Analysis (Bioheat Equation):
- The Pennes bioheat equation was simplified (ignoring blood/fluid perfusion) and solved using the forward finite difference method to estimate temperature rise (ΔT) due to light absorption (source term φµ_a).
Magnetic Field Forward Modeling:
- Extracellular magnetic fields (B) and electric potential (φ) were calculated from the simulated transmembrane and axial currents, initially assuming an unbounded homogenous volume conductor.
- A scaling factor was applied to the results to approximate the effects of conductivity anisotropy, the boundary between the brain slice and the sensor, and the surrounding fluid.

Commercial Applications

Quantum Sensing (Diamond): Driving requirements for next-generation NV diamond sensors, specifically demanding high sensitivity (sub-pT/Hz^1/2) and high spatial resolution for biological applications.
In Vitro Neural Imaging: Development of wide-field NV magnetometry systems for non-invasive, stain-free 2D functional imaging of neural networks in brain slices or cultured cells.
Optogenetics Research Tools: Providing simulation frameworks to optimize light delivery parameters (irradiance, duration, frequency) and select appropriate ChR2 mutants (e.g., ChETA, ChIEF) for achieving maximal neural synchronization and spiking efficiency.
Bio-Magnetometry: Establishing the expected signal strength benchmarks (0.1-0.5 nT) for fast APs in white matter pathways, guiding the design of sensor setups for detecting weak biological magnetic fields.
Computational Neuroscience: Utilizing detailed biophysical models (cable equation, ChR2 kinetics) to predict electromagnetic fields generated by complex neuronal morphologies, aiding in the interpretation of experimental NV data.

View Original Abstract

Magnetometry based on nitrogen-vacancy (NV) centers in diamond is a novel technique capable of measuring magnetic fields with high sensitivity and high spatial resolution. With the further advancements of these sensors, they may open up novel approaches for the 2D imaging of neural signals in vitro . In the present study, we investigate the feasibility of NV-based imaging by numerically simulating the magnetic signal from the auditory pathway of a rodent brainstem slice (ventral cochlear nucleus, VCN, to the medial trapezoid body, MNTB) as stimulated by both electric and optic stimulation. The resulting signal from these two stimulation methods are evaluated and compared. A realistic pathway model was created based on published data of the neural morphologies and channel dynamics of the globular bushy cells in the VCN and their axonal projections to the principal cells in the MNTB. The pathway dynamics in response to optic and electric stimulation and the emitted magnetic fields were estimated using the cable equation. For simulating the optic stimulation, the light distribution in brain tissue was numerically estimated and used to model the optogenetic neural excitation based on a four state channelrhodopsin-2 (ChR2) model. The corresponding heating was also estimated, using the bio-heat equation and was found to be low (&lt;2°C) even at excessively strong optic signals. A peak magnetic field strength of ∼0.5 and ∼0.1 nT was calculated from the auditory brainstem pathway after electrical and optical stimulation, respectively. By increasing the stimulating light intensity four-fold (far exceeding commonly used intensities) the peak magnetic signal strength only increased to 0.2 nT. Thus, while optogenetic stimulation would be favorable to avoid artefacts in the recordings, electric stimulation achieves higher peak fields. The present simulation study predicts that high-resolution magnetic imaging of the action potentials traveling along the auditory brainstem pathway will only be possible for next generation NV sensors. However, the existing sensors already have sufficient sensitivity to support the magnetic sensing of cumulated neural signals sampled from larger parts of the pathway, which might be a promising intermediate step toward further maturing this novel technology.

Tech Support

Original Source

DOI: https://doi.org/10.3389/fnins.2021.643614

References

2009 - Diamonds with a high density of nitrogen-vacancy centers for magnetometry applications. [Crossref]
2010 - Temperature dependence of the nitrogen-vacancy magnetic resonance in diamond. [Crossref]
2013 - Light scattering properties vary across different regions of the adult mouse brain. [Crossref]
2007 - An optical neural interface: in vivo control of rodent motor cortex with integrated fiberoptic and optogenetic technology. [Crossref]
2013 - Theoretical principles underlying optical stimulation of myelinated axons expressing channelrhodopsin-2. [Crossref]
2008 - Nanoscale imaging magnetometry with diamond spins under ambient conditions. [Crossref]
2009 - Ultralong spin coherence time in isotopically engineered diamond. [Crossref]
1995 - Pre- and postsynaptic glutamate receptors at a giant excitatory synapse in rat auditory brainstem slices. [Crossref]
2020 - Sensitivity optimization for NV-diamond magnetometry. [Crossref]
2016 - Optical magnetic detection of single-neuron action potentials using quantum defects in diamond. [Crossref]