Voltage Maps Expose the Shifting Electrical Terrain That Breeds Heart Rhythm Chaos

Every year, cardiac arrhythmias contribute to millions of deaths worldwide, yet one of the most fundamental questions in electrophysiology remains stubbornly difficult to answer: what exactly is happening in the heart muscle at the moment a deadly rhythm takes hold? A new study from researchers at the Federal University of ABC in Brazil offers an unusually detailed answer. By mapping the electrical voltage across the surfaces of isolated rabbit hearts, the team has captured, beat by beat, how the so-called arrhythmogenic substrate—the tissue landscape that permits dangerous rhythms to start and persist—behaves in ways that are far more dynamic than conventional clinical mapping assumes. The work, published in the Annals of Biomedical Engineering, establishes the first standardized epicardial voltage reference values for the rabbit heart and reveals that some low-voltage regions flicker in and out of existence with the rhythm itself, while others remain stubbornly fixed.

The clinical stakes are considerable. Cardiovascular diseases claimed roughly 19.8 million lives in 2022, about 32 percent of all global deaths, and arrhythmias sit at the dangerous intersection of stroke, heart failure, and sudden cardiac death. Electrophysiologists routinely rely on electroanatomical voltage mapping to find the scarred or electrically remodeled tissue that harbors these rhythms. The technique works by measuring peak-to-peak electrogram amplitudes across the heart and flagging regions that fall below established cut-off values as abnormal. But those cut-off values are anything but universal. They differ between the atria and the ventricles, between unipolar and bipolar recording configurations, between mapping technologies, and—critically for the research community—between species. Thresholds have been published for human, swine, ovine, canine, and rat hearts, yet despite decades of rabbit-based electrophysiology research, no standardized epicardial voltage values existed for the rabbit model. That gap is what the Brazilian team set out to close.

The experimental platform is a tour de force of controlled preparation. Nine New Zealand White rabbit hearts were extracted, heparinized, and connected to a Langendorff perfusion system, which pushes oxygenated, warmed Krebs–Henseleit solution backward through the aorta at a constant pressure of 70 to 80 millimeters of mercury, keeping the tissue alive and physiologically stable outside the body. Custom multi-electrode arrays—16 platinum electrodes arranged in a four-by-four grid on flexible PET substrates—were placed directly on the epicardium of the right atrium, the left atrium, and the anterior ventricular surface. The atrial arrays used 2-millimeter electrode spacing; the ventricular array spread its electrodes 4 millimeters apart to cover more tissue. Signals were acquired in a unipolar configuration at 4 kilohertz using Intan amplifiers and an Open Ephys acquisition board, all referenced to a tinned copper ring fixed around the aorta. To eliminate the motion artifacts that plague contracting hearts, the team perfused the muscle-relaxant blebbistatin, a selective myosin II inhibitor that uncouples contraction from electrical activity without dampening the electrical signals themselves.

With the hearts beating electrically but not mechanically, the researchers recorded baseline sinus rhythm from all three chambers—roughly 1,600 unipolar and 1,000 bipolar peak-to-peak amplitudes per chamber—after careful preprocessing that included detrending, 60-hertz notch filtering, and zero-phase Butterworth bandpass filtering between 0.5 and 250 hertz. Bipolar signals were derived by subtracting adjacent electrode pairs in horizontal, vertical, and diagonal orientations, with the best-defined morphology selected to ensure the derivation aligned with the local wavefront direction. The healthy-tissue threshold was then defined using the 95th percentile of the amplitude distribution, a non-parametric criterion widely used in clinical electroanatomical mapping that avoids assumptions of normality and resists distortion by extreme values. The resulting numbers are strikingly chamber-specific: unipolar thresholds came out at 7.4 millivolts for the right atrium, 8.9 millivolts for the left atrium, and 17.4 millivolts for the ventricle, while the corresponding bipolar thresholds were 4.1, 5.7, and 9.5 millivolts. Ventricular electrograms dwarfed atrial ones, a reflection of the ventricles’ greater muscle mass, more synchronized wavefronts, and different tissue architecture, and unipolar amplitudes systematically exceeded bipolar ones because unipolar electrodes sense a broader territory while bipolar measurements capture only the local voltage gradient.

Having established the baseline, the team turned to the heart of the question: what happens to these voltages when arrhythmia strikes? Using carbachol to modulate autonomic tone and pinacidil, an ATP-sensitive potassium channel opener, to shorten action potential duration, the researchers primed the hearts for arrhythmia and then induced atrial tachycardia, ventricular tachycardia, and ventricular fibrillation through burst pacing protocols delivered at 20 to 50 hertz. The amplitude collapses they documented were dramatic and rhythm-dependent. In one heart, ventricular peak-to-peak amplitudes fell from 31.0 millivolts during sinus rhythm to 8.3 millivolts during ventricular tachycardia, with a rank-biserial correlation of 1.000—the maximum possible effect size—and a median paired difference of 21.68 millivolts. During ventricular fibrillation, amplitudes crashed from 23.6 to 4.8 millivolts, again with an almost maximal effect size of 0.997. Cycle lengths shortened in parallel, from 278.7 to 184.4 milliseconds in tachycardia and from 460.2 to 145.6 milliseconds in fibrillation, consistent with the escalating frequency and disorganization of activation.

The atrial story proved more nuanced. During induced atrial tachycardia, the right atrium showed only a modest amplitude reduction, from 18.2 to 16.1 millivolts, with a moderate effect size of 0.348. The left atrium, by contrast, suffered a far steeper decline, from 28.2 to 9.9 millivolts, with an almost maximal effect size of 0.997 and a median paired difference of 15.94 millivolts. This chamber-specific asymmetry suggests that the left atrium is substantially more susceptible to amplitude suppression during rapid activation, a finding that resonates with clinical observations that the left atrium often dominates the substrate of atrial fibrillation. The researchers attribute the amplitude losses to impaired spatial coherence of electrical activity: as activation rates climb, wavefronts fragment, opposing vectors cancel one another, and the coordinated myocardial activation that produces large, clean electrograms disintegrates.

Perhaps the most visually compelling results came from the longitudinal voltage maps themselves. Because the arrays stayed fixed on the epicardium throughout each experiment, the team could compare maps of the identical tissue patch across four sequential states: baseline sinus rhythm, sinus rhythm immediately before the arrhythmia, the arrhythmia itself, and sinus rhythm after the arrhythmia terminated. During atrial tachycardia, this revealed a fascinating split personality. Certain electrodes—numbers 3, 4, and 15 on the right atrial array—showed low voltage persistently across every sinus rhythm map and during the arrhythmia, marking them as rhythm-independent, structurally suspect regions. But a second cluster of electrodes—1, 5, 9, 13, and 14—dropped below threshold only during the tachycardia and recovered fully once sinus rhythm returned. Low-voltage area coverage on that array jumped from zero percent at baseline to 18.75 percent before the arrhythmia, 50 percent during it, and back to 18.75 percent afterward. Those transient zones, the researchers concluded, represent a functional substrate: rate-dependent electrophysiological behavior that appears and disappears with the rhythm rather than fixed tissue damage.

The ventricular fibrillation maps told a fundamentally different story. What began as a homogeneous ventricular map—every electrode comfortably above the 17.4-millivolt healthy threshold—collapsed into near-universal low voltage once fibrillation began, with global amplitudes falling to 0.7 millivolts and every electrode dropping below 4 millivolts. Crucially, the suppression did not resolve when sinus rhythm returned. Post-arrhythmia maps showed only partial recovery, with all electrodes still below threshold and the same electrodes that had been most severely suppressed—89, 93, and 94—remaining under 2 millivolts. The spatial persistence of these low-voltage patterns across every time point after their first appearance suggests a substrate behavior compatible with sustained disruption of ventricular electrical organization, a pattern the authors interpret as predominantly structural in character, though they are careful to note that histological validation would be needed to confirm true tissue remodeling.

The authors are candid about the limitations that frame these conclusions. The three arrhythmias studied involve distinct mechanisms in different chambers, so their voltage patterns cannot be directly compared as evidence of a common arrhythmogenic pathway. The arrhythmia recordings were obtained in the continuous presence of blebbistatin, carbachol, and pinacidil, any of which could influence conduction, repolarization, and electrogram amplitudes. And the thresholds themselves are specific to the custom electrode geometry, spacing, signal-processing pipeline, and ex vivo preparation used here—they are platform-specific reference values, not universal biological boundaries, and would need validation before extrapolation to other rabbit mapping systems or in vivo settings. Within those bounds, however, the study delivers something the field has lacked: a quantitative, chamber-specific framework for interpreting voltage maps in the most widely used small-animal model of cardiac electrophysiology.

The broader implication is that voltage mapping, long treated as a static snapshot of scar and fibrosis, is actually a window into a living, breathing electrical landscape. The degree of amplitude suppression tracks the degree of electrophysiological organization, rising progressively from atrial tachycardia through ventricular tachycardia to fibrillation, and the spatial behavior of low-voltage regions distinguishes functional, rhythm-dependent abnormalities from persistent, potentially structural ones. For researchers using the isolated rabbit heart to probe conduction heterogeneity, electrophysiological remodeling, and the mechanisms that initiate and sustain lethal rhythms, the new reference values and the longitudinal mapping approach provide both a calibration standard and a methodological template. The next step, the authors suggest, is to pair these electrical maps with complementary structural validation, so that the flickering low-voltage zones captured on the epicardial surface can be traced to their ultimate biological causes—and, perhaps one day, to better-targeted therapies in human patients.

Subject of Research: Epicardial voltage mapping of arrhythmogenic substrate dynamics in Langendorff-perfused isolated rabbit hearts

Article Title: Peak-to-Peak Voltage Mapping Reveals Dynamic Arrhythmogenic Substrate Behavior in Isolated Rabbit Hearts

Article References: Peak-to-Peak Voltage Mapping Reveals Dynamic Arrhythmogenic Substrate Behavior in Isolated Rabbit Hearts. (n.d.). https://doi.org/10.1007/s10439-026-04382-5

Image Credits: AI Generated

DOI: 10.1007/s10439-026-04382-5

Keywords: cardiac electrophysiology, voltage mapping, arrhythmogenic substrate, rabbit heart model, Langendorff perfusion, atrial tachycardia, ventricular tachycardia, ventricular fibrillation, electrograms, low-voltage areas, multi-electrode arrays, sudden cardiac death