Skip to main content

m27 · RF ablation biophysics and lesion metrics (power, force, time, impedance drop)

Warning

Education only. Not for diagnosis, treatment or any clinical decision.

Note

Draft, stage 2 of 6. The derivations are written, and each equation is checked by tests in tests/test_m27_rf.py. Demos, code lab, questions and the remaining sections follow in later stages (plan: docs/outlines/m27-outline.md). Nothing here has had expert review yet. This module explains physics; it gives no power, force, time or index targets for patients.

In 60 seconds

Radiofrequency (RF) ablation heats tissue with electric current. The generator drives current from the catheter tip to a large patch on the skin. Near the small tip the current is crowded into a small volume, so that is where the tissue heats; half of the heat is made within one electrode radius of the tip surface. Heat then spreads outward by conduction and is carried away by blood. Tissue that stays hot enough for long enough is injured for good and stops conducting. Power, contact force, time and the fall in impedance are what a mapping system can show during an application. Each is an input to, or a clue about, the lesion; none of them measures it.

From the field

Planned in the outline (docs/outlines/m27-outline.md); written in a later stage.

Derivations

Symbols used throughout. \(I\) current (A). \(\sigma\) electrical conductivity of tissue (S/m). \(r_0\) electrode radius (m). \(r\) distance from the centre of the electrode (m). \(J\) current density (A/m²). \(q\) power density, the heat made per unit volume (W/m³). \(T\) temperature (°C), \(T_b\) blood and body temperature. \(k\) thermal conductivity (W/(m·K)). \(\rho\) density (kg/m³) and \(c\) specific heat (J/(kg·K)). \(F\) contact force and \(P\) generator power. \(t\) time (s).

Where the current goes and where the heat is made

At the frequencies used for RF ablation, 300 kHz to 1 MHz, tissue can be treated as purely resistive because displacement currents are negligible, so models solve a quasi-static (Laplace) problem for the voltage (Berjano 2006). Model the tip as a sphere of radius \(r_0\) in a large, uniform volume of tissue, with the return patch far away. The current leaves the sphere evenly in all directions, and the same current \(I\) crosses every sphere of radius \(r\), so the current density is

\[ J(r) = \frac{I}{4\pi r^2}, \qquad r \ge r_0 \tag{1}\]

Resistive (Joule) heating per unit volume is \(J^2/\sigma\):

\[ q(r) = \frac{J^2}{\sigma} = \frac{I^2}{16\pi^2\sigma\, r^4} \tag{2}\]

Heating falls as the fourth power of distance. Doubling the distance from the centre cuts it sixteen-fold. Integrating \(q\) over the shell between \(r_0\) and \(r\) gives the share of all tissue heating made inside radius \(r\):

\[ \frac{P(r_0 < r' < r)}{P_{\text{total}}} = 1 - \frac{r_0}{r} \tag{3}\]

Half of the heat is made between \(r_0\) and \(2r_0\), that is, within one electrode radius of the tip surface. Tissue farther away is heated mostly by conduction from this thin shell.

The same integral gives the resistance the tissue presents to the generator. Adding thin shells in series, each of thickness \(dr\) and area \(4\pi r^2\):

\[ R_s = \int_{r_0}^{\infty} \frac{dr}{4\pi r^2 \sigma} = \frac{1}{4\pi\sigma r_0} \quad\text{(sphere)}, \qquad R_s = \frac{1}{2\pi\sigma r_0} \quad\text{(hemisphere on an insulating flat face)} \tag{4}\]

The hemisphere carries the same current through half the area at every radius, so its resistance doubles. This spreading resistance near the tip dominates the tissue part of the circuit: most of the voltage drop, and most of the heating, happen close to the electrode.

A real tip is not buried in uniform tissue. Part of it touches the wall and part sits in blood, which has a different conductivity, so the current divides between the two; how much reaches the wall depends on contact and on the two conductivities (values: TODO(source)).

Key takeaways

  • Current density falls as \(1/r^2\) and heating as \(1/r^4\); half the heat is made within one electrode radius of the tip surface.
  • The spreading resistance \(1/(4\pi\sigma r_0)\) near the tip dominates the tissue’s share of the circuit.

The heat balance: the bioheat equation

Temperature follows a balance of four terms. Pennes wrote it for perfused tissue (Pennes 1948):

\[ \rho c\, \frac{\partial T}{\partial t} = \nabla\cdot\left(k \nabla T\right) + q - \omega_b \rho_b c_b \left(T - T_b\right) \tag{5}\]

From left to right: heat stored, heat conducted, heat made by the current, and heat carried away by blood flowing through the tissue at perfusion rate \(\omega_b\) (1/s), with blood density \(\rho_b\) and specific heat \(c_b\). Cooling at the surfaces enters as boundary conditions: models represent blood flowing over the wall with a convective heat-transfer coefficient at the tissue–blood interface, and an internally cooled electrode with a fixed temperature (Berjano 2006).

A steady state you can solve by hand. Keep spherical symmetry, constant properties and no perfusion, and wait until nothing changes. With \(q\) from Equation 2, the equation becomes \(k\,r^{-2}\,(r^2 T')' = -q\). Integrating twice and requiring \(T \to T_b\) far away:

\[ T(r) = T_b + \frac{A}{r} - \frac{B}{r^2}, \qquad B = \frac{I^2}{32\pi^2 \sigma k} \tag{6}\]

The constant \(A\) is set by the condition at the tip, for example a fixed interface temperature \(T_e\). The \(A/r\) term is pure conduction from a hot sphere; the \(-B/r^2\) term adds the heat made inside the tissue. Haines and Watson measured a hyperbolic fall of tissue temperature with distance in an isolated-tissue model, close to a model of this kind (Haines and Watson 1989).

Where is the lesion edge? If injury happens wherever the tissue reaches a temperature \(T_L\), the lesion edge is the isotherm \(T(r_L) = T_L\). Keeping only the conduction term, with the interface held at \(T_e\):

\[ r_L = r_0\,\frac{T_e - T_b}{T_L - T_b} \tag{7}\]

so the lesion radius is proportional to the electrode radius. Haines and colleagues found exactly this proportionality when they held the interface temperature constant and varied the electrode radius (Haines et al. 1990).

How fast heat spreads. Without the source, the heat equation spreads a point of heat as a Gaussian, \(\exp\!\left(-x^2/(4\alpha t)\right)\), where \(\alpha = k/(\rho c)\) is the thermal diffusivity (m²/s). The distance at which the profile has fallen to \(1/e\) of its peak is

\[ \ell = \sqrt{4\alpha t} \tag{8}\]

Conduction is slow and grows only with the square root of time. This is why a lesion keeps growing for a while after the surface has heated, and why the depth of a lesion depends on duration as well as power.

Key takeaways

  • Storage = conduction + RF source − perfusion; cooling at the tip enters as a boundary condition.
  • At steady state \(T = T_b + A/r - B/r^2\); an isotherm lesion edge scales with electrode radius.
  • Heat spreads over a distance of about \(\sqrt{4\alpha t}\): depth needs time.

From temperature to injury

A threshold temperature. The simplest rule draws the lesion edge at one temperature. In RF ablation models 50 °C is the most frequently used isotherm, although 48 °C and 59 °C have also been used (Berjano 2006). One experimental basis: in isolated guinea-pig papillary muscle heated for 60 s, excitability was lost reversibly between 42.7 and 51.3 °C and irreversibly at 50 °C or more (Nath et al. 1993). Tissue between those limits can stop conducting for a while and then recover, which is one reason an acute map can differ from a later one.

A rate process. Injury depends on time as well as temperature; Moritz and Henriques showed this for skin burns (Moritz and Henriques 1947). Arrhenius models, which go back to Henriques’ work on burns (Berjano 2006), treat injury as a first-order reaction: the undamaged fraction \(C\) falls as \(dC/dt = -k_d(T)\,C\) with rate \(k_d = A\,e^{-E_a/(R T)}\), where \(A\) (1/s) and \(E_a\) (J/mol) are tissue constants, \(R\) is the gas constant and \(T\) is in kelvin. Integrating gives \(C(t)/C_0 = e^{-\Omega(t)}\) with

\[ \Omega(t) = \int_0^{t} A\, e^{-E_a / (R\, T(\tau))}\, d\tau \tag{9}\]

\(\Omega = 1\) leaves \(1/e\) of the tissue undamaged (63 % damaged) and is a common choice for the lesion edge. The rule has limits: at moderate hyperthermic temperatures, about 43–50 °C, Arrhenius models often overestimate early cell death (Pearce 2013). Values of \(A\) and \(E_a\) for myocardium: TODO(source).

Key takeaways

  • A threshold isotherm is simple; a 50 °C edge has an experimental basis in isolated muscle.
  • The Arrhenius integral counts time at temperature; \(\Omega = 1\) means 63 % damaged.
  • Both are models of injury, with known limits at moderate temperatures.

Impedance and why it falls

The generator sees one series circuit: the tip, the tissue and blood near it, the rest of the body and the dispersive patch. Only the part near the tip changes much during an application. For a conductivity that varies with distance, the series-shell sum from Equation 4 still holds:

\[ R_{\text{near}} = \int_{r_0}^{\infty} \frac{dr}{4\pi r^2\, \sigma\!\left(T(r)\right)}, \qquad \sigma(T) \approx \sigma_0 \left[1 + \alpha_\sigma \left(T - T_0\right)\right] \tag{10}\]

The second part is a linear model with a temperature coefficient \(\alpha_\sigma\). Most RF ablation models use \(\alpha_\sigma\) = +2 %/°C, a value measured below 42 °C (Berjano 2006); some use other relations. Tissue conducts better as it warms, so \(R_{\text{near}}\) falls as the heated zone grows, and the shells nearest the tip count most because of the \(1/r^2\) weight. This is the physics behind the impedance drop shown during an application. In an ex vivo model, the drop in locally measured impedance correlated with lesion size (Kawano et al. 2022).

Key takeaways

  • Impedance falls because heated tissue conducts better, mostly in the shells next to the tip.
  • The drop is an indirect sign of heating near the tip, not a measure of lesion depth.

Lesion metrics

A mapping system can integrate what it measures during an application of length \(t_a\):

\[ \text{FTI} = \int_0^{t_a} F(t)\, dt, \qquad E = \int_0^{t_a} P(t)\, dt \tag{11}\]

The force–time integral (FTI, usually in g·s) counts how hard and how long the tip pressed. In a bench model with a moving tissue contact, FTI tracked lesion size at constant power (Shah et al. 2010). Delivered energy \(E\) (J) counts generator power, not power in the tissue.

Weighted indices combine force, power and time in one number; one such index uses contact force, time and power in a weighted formula (Das et al. 2017). We do not reproduce commercial formulas. A generic form is a monotone function \(g\) of a weighted integral:

\[ \mathcal{I} = g\!\left(\int_0^{t_a} F(t)^{a}\, P(t)^{b}\, dt\right) \tag{12}\]

with exponents \(a\), \(b\) and the function \(g\) fitted to lesion data. Every such metric leaves things out that the derivations above say matter: how the current divides between wall and blood, how much the tip is cooled, perfusion, wall thickness, and the time-at-temperature history. Two applications with the same FTI can make different lesions (the tests show this for a steady and a pulsing force trace).

Key takeaways

  • FTI and energy are integrals of what the system measures, not of what the tissue experiences.
  • Weighted indices are fitted surrogates; their weights hide model assumptions.

Clinical correlate

NotePending clinical review

Planned in the outline. Requires sign-off by an EP physician.

Visual summary

Planned in the outline.

How experts think

Planned in the outline.

Cheat sheet

Planned in the outline.

Interview questions

Planned in the outline.

Code lab

Planned in the outline.

Self-check

Planned in the outline.

Literature figures

Planned in the outline.

Further reading

Planned in the outline.

References

Berjano, Enrique J. 2006. “Theoretical Modeling for Radiofrequency Ablation: State-of-the-Art and Challenges for the Future.” BioMedical Engineering OnLine 5: 24. https://doi.org/10.1186/1475-925X-5-24.
Das, Moloy, Jonathan J Loveday, Gareth J Wynn, et al. 2017. “Ablation Index, a Novel Marker of Ablation Lesion Quality: Prediction of Pulmonary Vein Reconnection at Repeat Electrophysiology Study and Regional Differences in Target Values.” Europace 19 (5): 775–83. https://doi.org/10.1093/europace/euw105.
Haines, D E, and D D Watson. 1989. “Tissue Heating During Radiofrequency Catheter Ablation: A Thermodynamic Model and Observations in Isolated Perfused and Superfused Canine Right Ventricular Free Wall.” Pacing and Clinical Electrophysiology 12 (6): 962–76. https://doi.org/10.1111/j.1540-8159.1989.tb05034.x.
Haines, D E, D D Watson, and A F Verow. 1990. “Electrode Radius Predicts Lesion Radius During Radiofrequency Energy Heating. Validation of a Proposed Thermodynamic Model.” Circulation Research 67 (1): 124–29. https://doi.org/10.1161/01.res.67.1.124.
Kawano, Daisuke, Hitoshi Mori, Kenta Tsutsui, et al. 2022. “Time Dependency in the Radiofrequency Lesion Formation for a Local Impedance Guided Catheter in an Ex Vivo Experimental Model.” Journal of Arrhythmia 38 (6): 1080–87. https://doi.org/10.1002/joa3.12789.
Moritz, A R, and F C Henriques. 1947. “Studies of Thermal Injury: II. The Relative Importance of Time and Surface Temperature in the Causation of Cutaneous Burns.” The American Journal of Pathology 23 (5): 695–720.
Nath, S, C Lynch, J G Whayne, et al. 1993. “Cellular Electrophysiological Effects of Hyperthermia on Isolated Guinea Pig Papillary Muscle. Implications for Catheter Ablation.” Circulation 88 (4 Pt 1): 1826–31. https://doi.org/10.1161/01.cir.88.4.1826.
Pearce, John A. 2013. “Comparative Analysis of Mathematical Models of Cell Death and Thermal Damage Processes.” International Journal of Hyperthermia 29 (4): 262–80. https://doi.org/10.3109/02656736.2013.786140.
Pennes, H H. 1948. “Analysis of Tissue and Arterial Blood Temperatures in the Resting Human Forearm.” Journal of Applied Physiology 1 (2): 93–122. https://doi.org/10.1152/jappl.1948.1.2.93.
Shah, Dipen C, Hendrik Lambert, Hiroshi Nakagawa, et al. 2010. “Area Under the Real-Time Contact Force Curve (Force-Time Integral) Predicts Radiofrequency Lesion Size in an in Vitro Contractile Model.” Journal of Cardiovascular Electrophysiology 21 (9): 1038–43. https://doi.org/10.1111/j.1540-8167.2010.01750.x.