⚡ electrical engineering

A Lumped RC Equivalent Circuit Model of Head Tissues in sub-MHz Frequency Regimes

This paper presents a computationally efficient lumped RC equivalent circuit model that accurately reproduces the frequency-dependent electrical behavior of a three-layer spherical head up to 50 kHz, offering a viable alternative to costly numerical simulations for the design and evaluation of neuro-sensing and neuro-stimulation systems.

Original authors: Angelo Faccia, Ermanno Citraro, Francesco P. Andriulli

Published 2026-05-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Angelo Faccia, Ermanno Citraro, Francesco P. Andriulli

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your head is like a three-layered onion: the inner core is the brain, the middle layer is the skull, and the outer skin is the scalp. Now, imagine that inside this onion, tiny electrical signals (like neurons firing) are trying to send a message to the surface so we can measure them with electrodes on your skin.

For a long time, scientists have tried to predict exactly how these signals travel through the onion. The traditional way is like using a super-complex 3D map (called numerical methods like FEM). It's incredibly accurate, but it's also heavy, slow, and requires a massive computer to run. It's like trying to navigate a city by calculating the exact path of every single car, pedestrian, and pigeon in real-time. It works, but it's too slow for quick tests or for building devices that need to react instantly.

What this paper does:
The authors created a simple, lightweight "circuit map" (a lumped RC model) that acts like a shortcut. Instead of mapping every single cell, they built a small electrical circuit that mimics the behavior of the whole head. Think of it like replacing that massive 3D city map with a simple subway diagram: it doesn't show every street, but it tells you exactly how long the trip takes and how the train moves between stations.

Key Features of their "Subway Map":

The Onion Layers as Electrical Parts:
They turned the brain, skull, and scalp into simple electrical components:
- Resistors (R): These represent how hard it is for electricity to flow through the tissue (like a narrow pipe).
- Capacitors (C): These represent the tissue's ability to store a bit of electrical charge (like a sponge holding water).
- The Twist: Most old models treated the tissue like a simple pipe (just resistance). This paper realized that biological tissues are "spongy" and change their behavior depending on how fast the signal is moving (frequency). So, they added frequency-dependent parts. It's like realizing that a sponge gets harder to squeeze the faster you try to push water through it.
The "Dipole" (The Signal Source):
Inside the brain, the signal starts at a specific point (like a tiny battery). The authors realized that if this battery is right in the center, the signal spreads out evenly. But if it's closer to the edge (the skull), the signal behaves differently.
- The Analogy: Imagine dropping a stone in the middle of a pond versus dropping it right next to the edge. The ripples hit the shore differently. Their circuit model has a special "knob" that adjusts the resistance based on exactly where the signal starts, ensuring the prediction stays accurate even if the signal source is off-center.
Why the "Sponge" (Capacitance) Matters:
The paper tested what happens if you ignore the "sponge" effect (the capacitors) and just use the "pipe" (resistors).
- The Result: If you ignore the sponge, your prediction is wildly wrong. At higher speeds (frequencies up to 50 kHz), ignoring the capacitive effect made the predicted signal 160% too strong. It's like trying to predict how fast a car goes by ignoring air resistance; you'd think it's much faster than it actually is. The paper proves you must include these capacitive effects to get a realistic answer.

The Bottom Line:
The authors built a compact, fast, and accurate electrical circuit that acts as a stand-in for the complex human head.

It's fast: It runs on standard circuit simulators (like SPICE) instead of needing a supercomputer.
It's accurate: It matches the complex, heavy-duty math models almost perfectly across a wide range of speeds (from 10 Hz to 50 kHz).
It's smart: It accounts for the fact that head tissues change their electrical properties based on speed (dispersion) and that they store charge (capacitance).

What they claim (and what they don't):
The paper claims this model is perfect for designing and testing neuro-sensing and neuro-stimulation systems quickly. It allows engineers to prototype devices and simulate how they will work in real-time without waiting hours for a computer to crunch the numbers.

Note: The paper focuses strictly on the mathematical model and its validation against existing theories. It does not claim to cure diseases, diagnose patients, or be used directly in a hospital setting right now; rather, it provides the tool that engineers can use to build better medical devices in the future.

Technical Summary: A Lumped RC Equivalent Circuit Model of Head Tissues in sub-MHz Frequency Regimes

Problem Statement
Accurate modeling of electric potential and current distribution in head tissues is essential for the design and evaluation of neuro-sensing (e.g., EEG) and neuro-stimulation (e.g., tDCS, tACS, DBS) systems operating in the sub-megahertz frequency range. While numerical methods like the Finite Element Method (FEM) and Boundary Element Method (BEM) can handle complex geometries, their high computational cost limits their utility for rapid prototyping, real-time simulations, and direct integration into circuit-level design environments. Existing circuit-based approaches, such as Resistor Mesh Models (RMM), often rely on fine volumetric discretization, resulting in non-negligible computational demands, and frequently fail to account for the capacitive properties and frequency-dependent dispersion of biological tissues. There is a need for a compact, lumped-parameter equivalent circuit that maintains accuracy across a broad frequency spectrum while incorporating tissue dispersion and displacement current effects.

Methodology
The authors propose a lumped RC equivalent circuit model derived from a canonical three-layer spherical head geometry (scalp, skull, and brain) surrounded by an infinite air layer. The model operates within the electro-quasi-static (EQS) regime, utilizing complex frequency-dependent conductivity ( $\sigma_c(\omega) = \sigma(\omega) + i\omega\varepsilon(\omega)$ ) to capture both conductive and capacitive responses.

Circuit Topology: The model employs a minimal set of impedance elements. Each tissue layer is represented by two coupled branches: a radial impedance ( $Z^r_i$ ) modeling through-layer current flow and a tangential impedance ( $Z^\theta_i$ ) modeling intra-layer return currents. This dual-branch structure accounts for the fact that not all field lines traverse the entire volume; some current propagates within a shell before returning.
Boundary Conditions: Kirchhoff's Current Law (KCL) is enforced at interfacial nodes to satisfy the continuity of the normal component of total current density, naturally reproducing surface charge accumulation due to permittivity discontinuities.
Parameterization:
- Geometric Variability: The ratio between radial and tangential impedances is governed by a dimensionless geometric coefficient ( $\gamma_i$ ), which is a function of layer radii and thicknesses. Polynomial fits were derived to allow the model to adapt to variations in skull thickness and brain-scalp ratios.
- Dipole Eccentricity: To handle off-center neural sources, the brain's radial impedance is split into two asymmetric branches ( $Z^{r_{up}}_{brain}$ and $Z^{r_{down}}_{brain}$ ), scaled by an asymmetry parameter $\alpha(\eta)$ dependent on the dipole's radial position.
- Frequency Dependence: The model incorporates frequency-dependent conductivity and permittivity data (based on in vivo measurements by Wagner et al.) to model tissue dispersion.
Validation: The model was implemented in QSPICE and validated against a semi-analytical reference solution based on scalar spherical harmonics (SSH) expansion. Validation covered a frequency range of 10 Hz to 50 kHz, various dipole eccentricities, and different skull thicknesses.

Key Contributions

Compact Circuit Representation: The paper introduces a lumped-element circuit that reproduces the electrical behavior of a multi-layer spherical head using a minimal number of components, enabling integration into standard circuit simulators (SPICE).
Inclusion of Dispersive and Capacitive Effects: Unlike previous simplified models, this approach explicitly includes frequency-dependent conductivity and permittivity, as well as displacement currents, via complex impedance elements.
Geometric and Source Adaptability: The model parameters are designed to scale with anatomical variations (skull thickness, layer radii) and source eccentricity, making it robust for different physiological configurations.
Quantitative Assessment of Simplifications: The study quantitatively evaluates the errors introduced by neglecting displacement currents and frequency dispersion, demonstrating their critical role in model fidelity.

Results

Accuracy: The proposed lumped circuit model showed close agreement with the semi-analytical SSH reference solution across the 10 Hz–50 kHz range. The Mean Relative Frequency Error (MRFE) remained low, with minor deviations observed primarily at high dipole eccentricities where the source approaches the brain-skull interface.
Impact of Dispersion and Displacement Currents: Numerical analysis revealed that neglecting these effects leads to significant errors.
- A purely ohmic model (constant resistance, no capacitance) resulted in relative errors reaching approximately 160% at 50 kHz.
- Including frequency-dispersive conductivity alone reduced errors significantly but still produced discrepancies exceeding 10% from 600 Hz and up to 15% at 1.7 kHz.
- Only the inclusion of both frequency-dispersive resistance and capacitance (displacement currents) ensured physically consistent and accurate predictions.
Geometric Sensitivity: The error in scalp potential estimation was found to be more sensitive to dipole eccentricity than to variations in skull thickness, though both factors influenced the model's performance.

Significance and Claims
The authors claim that this work provides a practical, computationally efficient alternative to full volumetric numerical solvers for specific applications in neural sensing and stimulation. By offering a compact electrical representation compatible with circuit simulation environments, the model facilitates rapid design, signal-to-noise ratio assessment, and real-time adaptive tuning of stimulation parameters. The study emphasizes that accurate modeling in the sub-MHz regime requires the inclusion of both capacitive effects and tissue dispersion to avoid substantial overestimation of scalp potentials. While the current validation is limited to radially oriented dipoles in concentric, isotropic spherical layers, the framework establishes a foundation for future extensions to more complex geometries and anisotropic tissue properties.

More like this