HiPSC-CMs, iCell Cardiomyocytes² catalog # 01434 (Fujifilm Cellular Dynamics, Inc., Madison, WI), were handled according to the manufacturer’s instructions (fujifilmcdi, 2022). HiPSC‐CMs used in this study were derived from the hiPSC line, which was reprogrammed from fibroblast donor tissue isolated from an apparently healthy normal Caucasian female <18 years old (Ma et al., 2011; Feaster et al., 2021). Briefly, hiPSC-CMs, in a concentration of 110,000–115,000 cells per well, were plated on 96-well Nanofiber plates, catalog # 9602 (Nanofiber Solutions, Dublin, OH), coated with Matrigel substrate, catalog # 356230 (Corning Inc., Somerville, MA), in a 1:60 DMEM dilution, catalog # 30-2006 (ATCC, Manassas, VA). Cells were maintained using iCell Cardiomyocytes Maintenance Medium, catalog #M1003 (Fujifilm Cellular Dynamics, Inc.). Spontaneously beating, 100% confluent hiPSC-CM monolayers were used for PEF treatment testing on days 7–14 after plating. Hoechst-33342 (Ho) (2.25 µM), catalog #H3570 (Invitrogen, ThermoFisher Scientific, Waltham, MA) dye, labeling the nuclei of all cells, was used to assess monolayer confluency and integrity before pulsing (see Supplementary Figure S1). Two solutions were used during cell staining, PEF treatment, and imaging: iCell Cardiomyocytes Serum-Free Medium (iCM-SF), catalog #M1038, (Fujifilm Cellular Dynamics, Inc.), or modified Tyrode’s solution containing (in mM) 140 NaCl, 5 KCl, 2 CaCl₂, 2 MgCl₂, 5 HEPES, 10 Glucose, pH 7.4 with NaOH and electric conductivity of 1.8 S/m at room temperature, and 2.3 S/m at 37°C. PH and conductivity were measured with an S47 SevenMulti dual meter pH/conductivity (Mettler Toledo, Columbus, OH).

A pair of custom stainless-steel needle electrodes (.61 mm diameter, 1.7 mm distance center to center) were connected to an electric pulse generator, model FPG 1B50-1UL10 (FID GmbH, Germany) that was controlled with a digital delay generator, model 577-4C (Berkeley Nucleonics Corporation, San Rafael, CA). As in previous reports (Gudvangen et al., 2022), we used a 3D printer, model Anet A8 (Shenzhen Anet Technology Co., China), as an automated robotic arm for accurate positioning of the electrodes orthogonally to the cell monolayer; the tip of the electrodes was in contact with the bottom of the well. An electrode holder, connected to the 3D printer, was equipped with a spring system to minimize the pressure of the electrodes on the bottom of the plate. The 3D printer was programmed to move the electrodes to the center of each well with a 15–50 s delay, depending on the PEF parameter selection. The Anet A8 stage was heated to 50°C to adjust the pretreatment temperature of the Tyrode solution to 37.5°C ± 1.0°C. A multi-well plate holder (Olympus, Center Valley, PA) was secured, i.e., glued, to the 3D printer stage to avoid relative movements between electrodes and the multi-well plate. Pulse shape and amplitude were monitored with an oscilloscope (Tektronix, Beaverton, OR) connected to a 1:100 voltage probe, model P2501 (Owon Technology Inc., China) and an electric current probe, model TEK TCP 2020 (Tektronix).

A schematic of the experimental setup is reported in Supplementary Figure S2. The biphasic electric pulses were characterized by the following parameters: pulse repetition frequency (PRF), defined as the inverse of the pulse repetition period (PRP); number of biphasic pulses (p _#) delivered in a single train; phase amplitude (V_p); phase duration (t_p); and interphase delay (d_p) (Figure 1A). For each combination of PEF parameters, V_p was gradually increased until a clear measurable electroporation region was produced, while avoiding higher values of V_p that caused cell detachment and monolayer dissociation (see an example in Supplementary Figure S3). The interphase delay was maintained for all the PEF combinations at 1 µs.

Thermal changes due to PEF treatments were measured using a non-metallic optic STB probe, catalog # L-00-14500-01 (Advanced Energy Industries, Denver, CO) with a response time of .25 s, sampling rate of .02 s, and diameter of .5 mm. The STB probe was positioned adjacent and parallel to one of the electrodes (Supplementary Figure S2). To eliminate the influence of the STB probe on the PEF treatments, temperature measurements were carried out in cell-free plates separately from the experiments used for performing the PEF treatment analysis, as previously described (Arena et al., 2012). A representative temperature measure over time and the endpoints defined as maximum temperature increase (ΔT_max), 50% temperature recovery (ΔT_50%), and time-to-50%-recovery (Δt_50%) are shown in Figure 1B.

PEF-treated hiPSC-CM monolayers were stained with two cell-impermeable fluorescent probes: YO-PRO-1 Iodide (YP1) (5 µM) and Propidium Iodide (PI) (15 µM), catalog numbers Y3603 and P3566, respectively, (Invitrogen, ThermoFisher Scientific), at different times (Figure 2). 15 min before experimental cells were washed with 200 µl per well of dulbecco’s phosphate-buffered saline DPBS, catalog number 14190-144 (Gibco, ThermoFisher Scientific). DPBS was immediately replaced with 100 µl per well of modified Tyrode solution containing YP1. Stained cells by YP1 were imaged using confocal microscopy 30 min after PEF treatment. After imaging, the Tyrode solution was changed to iCM-SF, 100 µl per well, and plates were returned to a 37°C, 5% CO₂ cell culture incubator. To stain dead cells, iCM-SF was substituted with Tyrode containing PI 15 min prior imaging and imaged .5, 2, 4, 6, and 24 h after PEF treatment. The PI-stained area increased up to 2–4 h after PEF treatment, without further expansion at later time points (Supplementary Table S1).

For all the experiments described here, the 2–4 h timepoint was used to assess cell death areas. The laser scanning confocal microscope, model FluoView 3000 (Olympus) (Supplementary Figure S2) equipped with an environmental chamber (i.e., 37°C and 5% CO₂), model OKO-H301-OLY-IX3-SVR (Okolab stl, Italy), was used to collect fluorescent images from multiple wells using a 4X, NA/0.16 dry objective. The following settings and conditions were chosen: resonant one-way scanning mode with four times the average line and sequential scan per line (.067 µs/pixel, 520.025 ms/frame); YP1 was excited with a 488 nm laser and the emission of the dye was detected in the 488–540 nm range (varied to maximize signal for each plate); PI emission was excited with a 561 nm laser and detected in the 570–670 nm range (all plates).

To quantify the electroporated area, YP1-stained regions were analyzed with ImageJ software (NIH, Bethesda, MD) (Schneider et al., 2012). Images were converted to an 8-bit binary format (background threshold 12% ± 1%). Stacks of binary images were used as input to the Analyze Particle function that identified and quantified the area of the YP1-stained regions. The electrodes’ imprint area was quantified from sham exposures, i.e., .94 ± .18 mm², and subtracted from the electroporated area when needed.

PI-stained regions were analyzed using a custom MATLAB (MathWorks Inc., Natick, MA) code (Supplementary Appendix S1) that accounted for the gradual decrease in PI fluorescence at the border of the irreversibly electroporated region. Our code discretized each image in a matrix of 128 × 128 elements; for each element, the sum of red pixels with intensity above background (i.e., approximately 5%–15% of maximum intensity) was computed and normalized to 100%. The IRE area was obtained according to Eq. 1, multiplying the total number (N) of pixels (px) above a threshold value (th) by a conversion factor pixel² to mm² (c _f ). I R E   a r e a = c f ∑ i = 1 128 x 128 N i p x > t h (1)

In our analysis, to identify the margin of cell death area excluding regions where both YP1 and PI staining were present, th was set as 30% of the maximum density count (Supplementary Figure S4).

The external edge of the electroporation and cell death areas correspond to the minimum electric field, i.e., threshold, necessary to induce the effect. In this study, electroporation and cell death EFTs were identified by comparing the areas identified by YP1 and PI staining, respectively, to the electric field distribution from numerical simulations (Arena et al., 2012; Neal et al., 2014; Aycock et al., 2021; Liu et al., 2021; Aycock et al., 2022). The finite element analysis software Comsol Multiphysics 5.6 (COMSOL Inc., Stockholm, Sweden) was used to solve in static conditions the applied electric field map in the cell monolayer plane during pulsing. A 3D geometry of our pulse delivery system was constructed with equivalent dimensions for the electrodes and well used in the experimental setup (Supplementary Figure S5).

Each treatment was performed in a single well of a 96-well plate with a thickness of 1 mm and a radius of 3.5 mm. Stainless-steel electrodes (4∙10⁶ S/m) were placed perpendicular to the bottom of the well and inserted in a 2.75 mm thick layer of water with conductivity 2.3 S/m mimicking the modified Tyrode solution at 37°C. The geometry was surrounded by a cube of air (15 mm³ × 15 mm³ × 15 mm³, 0 S/m). The “finer” mesh setting, resulting in 166,212 elements, was used, with 5,910 elements for the electrodes (element volume ratio .0419) and 27,357 elements for Tyrode solution (element volume ratio 7∙10⁻³). One electrode was set to an electric potential equal to the phase amplitude of the PEF treatments, while the other electrode was set to 0 V. Electric field contours at varying magnitudes were created for each treatment, and the surface area within each contour was integrated with steps of 1 V/cm. Similarly to (Aycock et al., 2022), the curve fitting tool in MATLAB was used to fit a two-term exponential function to the resulting area versus electric field data. Finally, measured electroporation and cell death areas were used as inputs to this function to compute respective EFTs.

To estimate the temperature increase in adiabatic conditions, the adiabatic heating (AH, °C) was derived from the absorbed dose (AD, mJ/g) calculated according to (Ibey et al., 2009). Briefly, the following relationships were used: A H = A D 0.24 / 1000 (2) A D = σ E 2 ρ ⋅ τ ⋅ 10 6 (3) where τ = 2t_p is the pulse duration (s), E is the electric field in the solution (kV/cm) at the cell death threshold, ρ is the density of the solution (∼1 g/cm³), and σ is the conductivity of the solution (mS/cm).

Data are presented as mean ± standard error. Electroporation and cell death areas, as well as derived endpoints for a given set of PEF parameters were calculated as an average for n = 3–8 per group. In Figure 4, statistical analysis was performed using the Mann Whitney Test for unpaired data, significance was determined to be p < .05 before Bonferroni correction. Temperature measurements were calculated as an average for n = 3 per group.

We first evaluated the effects of PEF treatments (Figure 1A) in hiPSC-CMs. Robust electroporation and cell death of the hiPSC-CM monolayers was observed by fluorescence staining after PEF application (Figure 2). YP1 staining 30 min after treatment in response to trains of p _# = 300 delivered at PRF = 10, 100, and 1,000 Hz, V_p = 144, 236, and 472 V, and t_p = 1, 5, and 10 µs, resulted in a PEF parameter-dependent uptake (Figure 3A). In fact, the area of electroporation in hiPSC-CMs varied with changes in PEF parameters. The cell death areas indicated by delayed PI staining (Figure 3B) followed the same PEF parameter-dependent trend as areas indicated by YP1 (Figure 3A). However, cell death areas, labeled PI, were smaller than electroporated areas, labeled YP1, consistent with the expectation that cells located further from the electrodes, and therefore exposed to a lower electric field, would be more capable of repair (reversible electroporation). These results demonstrated that PEF treatments elicit detectable electroporation and cell death in an in vitro human cardiomyocyte model.

EFTs for electroporation and cell death of hiPSC-CMs were evaluated from YP1 and PI staining, respectively, for a wide range of PEF parameters. Pulses with t_p = 1, 5, and 10 µs were delivered in trains of p _# = 50, 100, and 300 at PRF = 10, 100, and 1,000 Hz. The EFT for cell death was inversely proportional to t_p and p _#, and directly proportional to the PRF (Figure 4; Table 1). For example, we found that for hiPSC-CMs an increase of t_p from 1 to 5 µs (p _# = 50, PRF = 1,000 Hz) produced a decrease of EFT for cell death from 3.15 to 1.56 kV/cm. Increasing the p _# in the train reduced the EFT for cell death, with values down to 2.23 kV/cm for p _# 300 (t_p = 1 µs, PRF = 1,000 Hz). The EFT for cell death increased from 2.86 to 3.15 kV/cm when the PRF increased from 10 to 1,000 Hz (t_p = 1 µs, p _# = 50). The EFT for cell electroporation was consistently lower than the EFT for cell death (Figure 4).

To assess the proportion of electroporated and dead cells, we computed the ratio in percent of the EFTs determined by YP1 and PI staining. The EFT ratio was between approximately 80%–100%, showing a larger population of dead cells with increasing PRF and t_p (Table 1). As this ratio approaches 100%, the electroporation and cell death areas reach a comparable surface, ideal for ablation treatments where reversible effects are not desired.

Reduction in the EFT for cell death and in reversible electroporation beyond the cell death region are only two of the criteria to be considered when designing ablation strategies. PEF can induce heating that should be minimized to avoid cell death by thermal damage. Since temperature increase due to Joule heating is proportional to the absorbed doses, energy delivery minimization should be considered when selecting PEF treatment parameters.

To evaluate the energy deposition by combinations of PEF parameters, we calculated the AD at the cell death EFT as described in (Ibey et al., 2009). Variation in PEF parameters produce significant changes in the AD that ranged from approximately 20–80 J/g (Table 1). For all PEF combinations, trains with lower p _# achieved cell killing at lower doses, i.e., 18.8–29.4 J/g for 50 p _# compared to 46.8–79.9 J/g for 300 p _#; for the same p# in the train, the reduction in t_p and PRF decreased the delivered dose due to the lower energy deposition by shorter and less frequent pulse applications (Figure 5A). High ADs might result in a non-negligible temperature increase that should be considered a factor when planning PEF treatments.

These results demonstrated that PEF parameters can be modulated to reduce the EFT for cell death, undesired reversible effects, as well as the energy deposition during treatment. However, the high values of AD reached for certain pulsing conditions indicated that temperature assessment is suggested during PEF ablation.

PEF-induced temperature changes (Figure 1B) were measured in proximity to one of the electrodes (Supplementary Figure S2). PEF treatments induced a ΔT_max that ranged between 3°C and 11°C (corresponding to an absolute increase in temperature from 37°C baseline to 40°C and 49°C) and a Δt_50% that ranged between 3.2 and 12.8 s. The highest ΔT_max measured was 11°C and the maximum Δt_50% 12.8 s when a train of p _# = 300, PRF = 100 Hz, t_p = 5 µs, and V_p = 236 V was applied. For the longest train tested, i.e., p _# = 300 at PRF = 10 Hz, the recovery time was less than 30 s (Table 1). In general, ΔT_max increased with p _#, and t_p, while PRF showed a variable trend with often a peak at 100 Hz, suggesting that the temperature probe, due to its response time, may underestimate the highest temperature increase for high repetition rates, e.g., f = 1,000 Hz. To estimate the maximum temperature increase possible independently of specific environmental and experimental factors, we also computed the AH at the cell death EFT, as described in (Ibey et al., 2009). For low 50 and 100 p _# the AH was moderate and ranged between 4°C and 12°C (Figure 5B; Table 1). However, for higher PEF doses, temperature values reached 20°C, likely resulting in some thermal damage (Davalos et al., 2005).

Our results demonstrated that this assay allows for the quantification of critical factors in the selection of PEF parameters for cardiac ablation treatments. In order to select the optimal parameters for PEF cardiac ablation among the combinations studied in this work, we ranked the PEF treatments with two multiple-criteria decision analysis methods. Both the Weight Sum Model (WSM; also known as weighted linear combination or simple additive weighting) and Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) are multi-criteria decision analysis methods used to rank the 27 various scenarios or “alternatives” consisting of multiple metrics or “criteria”. The WSM method uses a weighted sum to determine ranking, while TOPSIS uses L2-distance to measure distance to a best-case scenario of the reported criteria (Fishburn, 1967; Tzeng and Huang, 2011). Before processing our data with WSM and TOPSIS, weights to each of the critical factors considered were assigned arbitrarily considering relevance to clinical applications: 5 to the EFT for cell death, 3 to AH, 2 to AD, 2 to ratio electroporation/cell death EFTs, 1 to the total treatment time. Our analysis using both methods ranked the following treatments as the top two: i) p _# = 50, t_p = 10 µs, PRF = 100 Hz, and ii) p _# = 50, t_p = 10 µs, PRF = 10 Hz. These PEF treatments were characterized by an EFT for cell death equal to 1.07 and 1.00 kV/cm, an AH of 6.3°C and 5.6°C, an AD of 26.3 and 23.2 J/g, respectively, a ratio electroporation/cell death of 100% and 95%, and a total treatment time of .5 and 5 s (Table 1).

In this study, we established a robust high-throughput experimental protocol for efficient testing of a wide range of PEF parameters in a commercially available in vitro human cardiomyocyte model for cardiac ablation treatment optimization. PEF-based cardiac ablation devices have been gaining increasing interest for the treatment of AF in patients that do not respond to pharmaceuticals. However, the lack of preclinical standardized methods to optimize PEF parameter selection for safe and efficient patient therapy has slowed device development and regulatory decision-making. Previous studies relied primarily on animal models or various cell lines and reported a limited range of PEF parameters (Xie et al., 2015; Xie and Zemlin, 2016; Semenov et al., 2018; Azarov et al., 2019; Neuber et al., 2019; Heller et al., 2021; Baena-Montes et al., 2022). Here, we report on an assay that enables the quantification of electroporation and cell death in hiPSC-CM monolayers following the application of PEF treatments with varying pulse parameters. Combined with temperature measurements, this approach can facilitate optimal PEF parameter selection ensuring complete PEF-induced cardiac ablation while minimizing undesired secondary effects (e.g., reversible effects and thermal damage).

Cell electroporation and cell death by PEF are threshold phenomena for which reversible and irreversible cell permeabilization occur only if the applied electric field in a tissue reaches a certain value. While increasing V_p is the most direct way to increase the ablation volume, it also increases the chances of muscle contraction and temperature effects. Minimization of the EFT for cell death by tuning different pulse parameters enables achievement of the desired ablation volume while keeping the V_p low (Sano et al., 2018). Additionally, for ablation procedures, it is desirable to select PEF parameters able to reduce the volume of reversible electroporation that could cause stimulation outside the ablation-targeted region or stunning and consequent recurrence of AF. Finally, this parameter selection should guarantee that the temperature of the tissue remains at or below physiological ranges to avoid thermal damage.

In previous studies, needle electrodes producing a non-uniform electric field distribution have been used to identify the EFT of electroporation and cell death in cell monolayers and in 3D in vitro models (Arena et al., 2012; Sano et al., 2018; Aycock et al., 2022; Gudvangen et al., 2022). These studies demonstrated that the EFTs can be quantified by matching the area of the stained surface obtained experimentally to the electric field map obtained numerically. Using this approach, we quantified the electroporation and cell death EFTs of hiPSC-CMs for biphasic pulses with t_p = 1, 5, and 10 µs delivered in trains of p _# = 50, 100, and 300 at PRF = 10, 100, and 1,000 Hz. Consistent with a previous study on a different cell type (Gudvangen et al., 2022), we found that for hiPSC-CMs t_p was the PEF parameter that mostly influenced the EFT for both electroporation and cell death with an inversely proportional relationship. As the t_p increased by one order of magnitude, the EFTs decreased non-linearly approximately 3 fold (see Table 1 for absolute values). Conversely, increasing the number of pulses in the train from 50 to 300 produced a decrease in the electroporation and cell death EFTs of only 1.4–1.8 fold (see Table 1 for absolute values). Preliminary data extending our results to p _# = 500 resulted in plateauing of the EFTs (data not shown), suggesting that adding pulses to the train only produced additional heating without further widening of the electroporation and cell death areas, i.e., lowering of the EFTs. In the range of the pulse parameters selected, increase of the PRF by three orders of magnitude led to a moderate but significant increase of the EFTs.

When designing a PEF-based ablation treatment, the EFT for electroporation compared to cell death indicates possible transitory effects, such as action potential (AP) generation, or stunning near the ablation area. In our study, the minimum ratio of EFTs for electroporation and cell death was 84%, showing generally a larger population of dead cells than of reversibly electroporated cells.

Another consideration during the implementation of PEF ablation treatments is the temperature increase due to Joule heating. While it is often suggested that IRE ablation is a thermal damage-free approach, PEF can produce thermal effects if treatment parameters are not selected correctly. We monitored temperature changes throughout the experiments and measured a maximum temperature increase of 11°C and maximum recovery time of 13 s. However, these measurements are dependent on the specific experimental environment. To estimate the worst-case scenario for temperature increase, we computed the AH at the cell death EFT. AH calculations, in fact, provide the maximum possible temperature increase, since heat dissipation and heat losses are not allowed. For low pulse numbers in the train, the AH ranged between 4°C and 12°C, remaining below the threshold for instant cell death by thermal damage (Davalos et al., 2005). While the quantitative assessment of the actual thermal damage to PEF-exposed cells was outside the scope of this paper, for higher doses the AH reached 20°C, indicating that some degree of cell death by thermal damage, especially close to the electrodes where the electric field is higher, cannot be ruled out (Davalos et al., 2005; Garcia et al., 2011; Long et al., 2014).

In summary, our results demonstrated that this assay allows for the quantification of critical factors for the selection of PEF parameters for cardiac ablation treatments, such as the EFT for cell death, the ratio of EFT for electroporation/cell death, and the temperature increase due to PEF treatment. For example, a train of 300 pulses with t_p = 10 µs, PRF = 10 Hz, with a total treatment time of 30 s, resulted in the lowest EFT for cell death measured (.59 kV/cm), with a ratio for EFT of electroporation/cell death at 94.4%; however, the AD was significant (48.9 J/g) as was the temperature increase (4.3°C measured, 11.7°C calculated).

In fact, minimization of the EFT for cell death should not be the only indicator for the selection of PEF treatments. One possible criterion is the identification of the set of parameters that minimizes the EFT for cell death and maintains the ratio of electroporation/cell death close to 100%, while reducing the AD, the thermal increase, and the total treatment time. One way to do this is to rank PEF treatments with a multiple-criteria decision analysis. Our analysis, using both WSM and TOPSIS methods, indicted that the top two PEF parameter combinations that best matched the selection criterion above were p _# = 50, t_p = 10 µs, PRF = 10, and 100 Hz. These two combinations, with respect to the one selected only based on the minimum cell death EFT, showed a drastically reduced treatment time, (i.e., 6-, 60-fold lower), up to 5% higher ratio of electroporation/cell death EFTs, and an almost 50% reduction in temperature increase and AD, with only a modest 1.8-fold increase in EFT for cell death (see Table 1 for absolute values).

One limitation of our study is the relatively restricted range of PRF values investigated, due to the capabilities of our pulse generator. Extending our results to higher PRF values could cover common treatments in preclinical studies, such as H-FIRE, i.e., high-frequency IRE, approaches that deliver tens of pulses in the kHz–MHz range (Arena et al., 2011), and cutting-edge nanosecond PEF approaches (Gudvangen et al., 2022).

Cell detachment due to PEF application is often observed during electroporation experiments, especially in excitable cells with contractile properties. PEF can elicit APs by either direct effects on voltage gated channels or as a downstream effect of cell electroporation that leads to loss of resting membrane potential (Casciola et al., 2019). In cardiac cells, this ionic imbalance can lead to contraction and eventual detachment from the plate. To enhance cell adherence, and to avoid cell detachment and peeling of the monolayer, we used patterned nanofiber plates. We cannot rule out some degree of undesired membrane damage during cell contraction induced by PEF application, which occurred due to the limited elasticity of the plates used (Graybill et al., 2021). This could be one possible explanation for the unexpected gradual PI uptake we observed in our experiments. To mitigate this limitation, future work will investigate the impact of exposing hiPSC-CM monolayers cultured on more flexible substrates, such as PDMS (Herron et al., 2016), on PEF treatments.

While PFA has been mostly proposed for the treatment of AF, our human model is composed of multiple cardiac cell subtypes, (i.e., ventricular, atrial, nodal). Approximately 50%–80% of the population is ventricular and the remaining 20%–50% is a combination of nodal and atrial (Ma et al., 2011). To the best of our knowledge, limited information is available on different sensitivities of atrial and ventricular cardiomyocytes to PEF, with some studies suggesting that atria are more susceptible to electroporation than ventricles (Fedorov et al., 2008; Neven et al., 2014). Currently, it is not clear if this difference results from their decreased thickness and increased heterogeneity (macroscopic tissue properties), or from different membrane shape and structure (cellular properties). While this could be relevant in the assessment of tissue specificity to PEF, we would not expect significant changes in the general trends reported here.

Finally, our assay is based on a 2D in vitro model, while clinical applications involve 3D in vivo tissues. The quantitative results presented here might differ for in vivo treatments, but general trends regarding how the thresholds change with pulse parameters would likely be maintained (Gudvangen et al., 2022). Additional studies may be able to predict the ablated volume and corresponding EFTs for specific PEF treatments in 3D tissues based on the results from our 2D in vitro assay. To support translatability of the results obtained with this in vitro assay to in vivo models, we are currently implementing a computational, i.e., in silico, cardiac model that will be coupled with the EFTs identified in this work to predict specific ablation areas for combinations of PEF parameters. The in silico results will be validated with experiments on isolated perfused porcine hearts by histological assessment of ablated tissues, justifying the in vitro assay as a reliable surrogate for in vivo testing.