NeuraSense Labs

Exploring the electrical signatures of tissue.

Different tissues can respond differently to a very small alternating current. Our work investigates how measurements taken through a surgical instrument can be characterized and analyzed for future surgical-sensing applications.

Bioimpedance Instrument-integrated concept Research and development

01 The principle

Every tissue has an electrical character.

Impedance is how strongly a material opposes an alternating current. Pass a very small current through tissue, measure the voltage that results, and the ratio between the two is the tissue’s impedance.

It has two parts. The magnitude says how large the voltage is for a given current. The phase says how far the voltage wave lags behind the current wave. A pure resistor shows no lag; anything that stores charge, like a capacitor, does.

Frequency matters because of how tissue is built. Every cell is wrapped in a thin membrane that separates its conducting interior from the conducting fluid around it, so each membrane behaves like a tiny capacitor. At low frequencies the membranes block most of the current, which flows around the cells through the extracellular fluid. As the frequency rises, more of the current crosses the membranes and passes through the cells themselves. Impedance falls, and the phase shifts, across that transition.

At low frequency, current paths run through the fluid between the cells and do not enter them.

Low frequency Membranes block most of the current, which flows through the fluid between cells.

At high frequency, current paths also pass straight through the cells, crossing their membranes.

High frequency Current also crosses the membranes and passes through the cells.

Fig. 1 Schematic, not to scale. Current passes between contacts at the left and right edges.

What shapes a tissue’s spectrum

  • Cell size and packing

    Tightly packed cells leave narrow channels between them, so less current gets through at low frequencies.

  • Membrane integrity

    Intact membranes create the characteristic transition. Damaged membranes, for example after heating, weaken it.

  • Fluid content

    Water and dissolved ions carry the current. More fluid in and around the cells generally means lower impedance.

  • Fiber orientation

    Skeletal muscle conducts differently along and across its fibers, particularly at lower frequencies, so the reading depends on direction.

  • Fat

    Fat cells are large and filled with lipid, which conducts poorly, so fatty tissue tends to show high impedance.

02 Impedance explorer

Interactive model

Turn the frequency. Watch the response change.

The Cole model is a standard textbook description of how tissue-like materials respond across frequency. Two illustrative parameter sets show how a different structure changes both the size of the response and its timing.

Fig. 2 · Impedance explorerCole model · illustrative values

Static view

With JavaScript turned on, this figure adds a frequency control, live model readouts and the applied and resulting waveforms. Here it shows both models across the full frequency range.

Spectrum

  • Model A
  • Model B
Relative magnitude and phase of two illustrative Cole models across frequency. Frequency runs from 0.01 to 100 times the reference frequency on a log scale. Model A’s magnitude falls from 1.0 to about 0.2, and its phase reaches about minus 31 degrees near 2.8 times. Model B’s magnitude falls from about 0.64 to about 0.26, with a shallower phase minimum of about minus 17 degrees near 8 times. The two magnitudes cross near 1.9 times.

Cole arc

−X (reactance) against R (resistance)

Cole arcs of both models. Each traces a flattened semicircle from its low-frequency resistance on the right to its high-frequency resistance on the left. Model A’s arc is larger.

Small dots mark each tenfold step in frequency, which rises from right to left. The top of each arc sits at that model’s characteristic frequency.

Fig. 2 Illustrative textbook model (Cole). Parameter values are arbitrary; these are not NeuraSense measurements.

Model background and references

Model A
R0 1.00 · R∞ 0.20 · fc 1× · α 0.80
Model B
R0 0.64 · R∞ 0.24 · fc 4× · α 0.72
Key
R0, R∞: low- and high-frequency limits of this dispersion. fc: characteristic frequency. α: spread of relaxation times (1 is a single relaxation time).

What to look for

  1. Well below fc

    Current flows around the cells

    Magnitude is high and nearly flat, and phase is close to zero. Membranes block the current, so it travels through the fluid between cells and the tissue behaves almost like a simple resistor, R0.

  2. Around fc

    Membranes charge and discharge

    Magnitude falls fastest. The out-of-phase part of the response peaks at fc, the top of the Cole arc, and the phase angle is largest close by. Each cycle, the membranes charge and discharge like capacitors.

  3. Well above fc

    Current crosses the membranes

    Magnitude levels off toward R∞ and phase returns toward zero. The membranes no longer hold the current back, so it passes through the cell interiors as well as around them.

03 At the instrument

Concept image: the fine tips of steel surgical forceps against a black background, with thin lines of violet light.
AI concept visualization Instrument tips at the point of contact.

Sensing at the point of contact.

In the NeuraSense concept, the measurement is taken through the tips of a bipolar forceps, a standard surgical instrument that already grasps and holds tissue. The two tips serve as the two contacts for the test signal, so the measurement samples the region around the instrument tips, with the sampled volume shaped by electrode geometry and contact conditions.

That placement is the point. The aim is information where the surgeon is already working, without a separate probe or a pause to change instruments. It also defines the engineering problem: a forceps is held by hand, closes with varying force, works in a wet field and, in many procedures, delivers electrosurgical energy. Our measurement design work addresses each of these conditions.

Six engineering priorities for instrument-integrated sensing.

Our development priorities cover the contact, measurement environment and presentation of tissue information.

  1. 01 Contact and pressure

    Characterize contact and grip force

    Impedance depends on the contact area at the tips and on how firmly tissue is held, and both change from moment to moment. We characterize that dependence and look for measures that hold steady as grip varies.

  2. 02 Interface and calibration

    Account for the electrode–tissue interface

    Where metal meets tissue, electrode polarization adds an impedance of its own, strongest at low frequencies. Calibration against known reference loads characterizes the instrument’s contribution, and tracking how the tip surface changes with use keeps readings comparable over time.

  3. 03 Signal quality

    Maintain signal quality in a wet field

    Blood, irrigation fluid and thin films of liquid can open conduction paths between the tips that bypass the tissue. We characterize signal quality under the conditions surgery presents, as well as on the bench.

  4. 04 Tissue state

    Measure temperature and tissue-state effects

    Temperature, blood flow, hydration and earlier energy application all change tissue’s electrical properties. Characterization separates these effects from differences between tissue types.

  5. 05 Normal use

    Preserve normal instrument function

    Bipolar forceps grasp, dissect and often deliver electrosurgical energy, and sensing should leave that work unchanged. Energy delivery can overwhelm sensing signals, so the measurement design must establish when valid readings are available and exclude readings affected by interference.

  6. 06 Presentation

    Present information at the right moment

    Useful information is simple, arrives when it is needed and stays quiet when it has nothing to add. Presentation is a research area in its own right, and one that benefits from surgeons’ perspective. Ways to collaborate.

04 Instrument to information

Measure. Characterize. Evaluate.

Measurement, characterization and evaluation connect tissue response to interpretable information.

  1. 01

    Measure

    In
    Tissue in contact with the instrument tips
    Out
    Calibrated magnitude and phase at each frequency

    A small alternating test signal passes between the tips, and the response is recorded as magnitude and phase, one frequency at a time across a range of frequencies.

    Evaluation focus

    • Stable contact through each reading.
    • Calibration against known reference loads at the start and end of a session.
    • Noise, drift and interference within characterized limits.

  2. 02

    Characterize

    In
    Readings across many frequencies
    Out
    A few features that describe the spectrum’s shape

    The readings form a spectrum. Its shape is summarized in a small set of descriptive features, such as the parameters of a fitted Cole model, so that spectra can be compared fairly.

    Evaluation focus

    • Differences that are consistent across specimens, sessions and setups.
    • Every spectrum traceable to its specimen, setup and conditions.
    • Features that stay stable as contact and temperature vary.

  3. 03

    Evaluate

    In
    Features from new, held-out spectra
    Out
    Performance estimates on unseen specimens, with their uncertainty

    Analysis and classification methods are tested on data kept strictly apart from the data used to build them, to learn how well a spectrum indicates tissue type and where its limits lie.

    Evaluation focus

    • Training and evaluation data split by specimen.
    • Uncertainty reported with every estimate.
    • Results that hold when the work is repeated.

Then · Vision

The long-term aim is clear, timely tissue information at the point of contact that supports the surgeon’s judgment.

05 Evaluation

Evaluating performance in the intended setting.

Our evaluation priorities cover repeatability, instrument function, useful outputs and the intended surgical setting.

Read: Evaluating tissue-sensing data

Repeatable differences
Tissue differences that appear consistently across specimens, sessions and setups.
Stable calibration
Readings that stay comparable across instruments, sessions and days.
Instrument function
Forceps that grasp, handle and deliver energy as surgeons expect, with sensing added.
Useful outputs
Information that is clear, timely and quiet when it has nothing to add, supporting the surgeon’s judgment.
Intended surgical setting
Evaluation under the procedures, conditions and users the technology is intended for.

06 Continue

Explore our research approach.

See how we organize measurement, data provenance and repeatable evaluation across our development work.

Research overview Contact us