Waveform shape emerges as a marker of local cortical kinetics
A new bioRxiv preprint argues that the contour of a cortical oscillation is not just a visual artifact of signal processing. It carries information about how local excitatory and inhibitory circuits behave, and about how those circuits interact with other brain regions. By combining a modified excitatory-inhibitory Wilson-Cowan model with human stereo-EEG and macaque electrocorticography, the investigators tied waveform morphology to synaptic kinetics, phase-dependent responsiveness, and multiple measures of interregional coupling.
The central idea is straightforward but biologically rich. Rhythms generated by cortical tissue do not oscillate as perfect sine waves, because real neural populations are shaped by asymmetries in excitation and inhibition, differences in rise and decay times, and state-dependent coupling with distant areas. Those details leave a fingerprint in the waveform itself. In this study, the fingerprint was most tightly linked to local irreversibility, while other aspects of the signal tracked broader network interaction measures.
One finding stood out immediately: waveform asymmetry tracked local circuit dynamics more strongly than interregional effects in the human recordings.
Synaptic rise and decay kinetics altered morphology and responsiveness
The modeling work focused on a modified excitatory-inhibitory Wilson-Cowan framework, a classic tool for describing population-level neural activity. Within that model, synaptic rise and decay kinetics shaped both waveform geometry and phase-dependent responsiveness. In practical terms, the same nominal oscillation can look and behave differently depending on how quickly excitation ramps up, how long inhibition persists, and where in the cycle a perturbation lands.
This matters because cortical oscillations are often interpreted through frequency alone. A rhythm at alpha frequency, for example, is commonly summarized by its mean instantaneous rate. Here, however, the shape of the waveform and the timing of perturbations relative to the cycle also mattered. Local kinetics and coupling symmetry jointly influenced interaction measures and temporal irreversibility. Phase-locked perturbations changed imaginary coherency and bivariate irreversibility even when coupling strengths stayed fixed, which suggests that timing within the cycle can alter measurable network dynamics without any change in the nominal connection magnitude.
That result points to an important mechanistic distinction in neural systems. Connectivity is not only a matter of how strongly regions are linked; it is also a matter of when a population is most excitable, how long it remains responsive, and whether the shape of the oscillation favors forward or backward propagation of influence.
Human stereo-EEG separated local from interregional signatures
The human component used stereo-EEG from 23 patients. In those data, waveform asymmetry associated most strongly with local irreversibility, whereas mean instantaneous alpha frequency showed stronger associations with interregional measures. The separation is conceptually useful. Waveform asymmetry appears to reflect the kinetics of the local generator, while frequency may be more sensitive to how one cortical site interacts with another over distance.
The authors did not present this as a generic relationship across all signals. Instead, the pattern depended on which feature was examined. When waveform morphology was the focus, the local circuit signature became more prominent. When alpha frequency was the focus, the network-level associations became stronger. The same recording can therefore contain at least two layers of information: one about the local neural machinery producing the rhythm, and another about the way that rhythm participates in cortico-cortical communication.
For clinicians and neuroscientists who work with invasive electrophysiology, the implication is concrete. A flattened summary of oscillatory activity can hide biologically relevant structure. A waveform that appears merely asymmetric may be carrying a readout of local excitatory-inhibitory balance, especially when paired with measures of irreversibility.
Macaque recordings preserved the frequency signal but not the local waveform link
Subdural macaque electrocorticography reproduced the frequency associations but not the local waveform association. That divergence is useful rather than disappointing, because it shows where the human result may depend on recording configuration, species-specific circuitry, spatial sampling, or the particular cortical scale accessed by stereo-EEG. The macaque data supported the broader link between frequency and interregional measures, while the human recordings exposed a stronger relationship between waveform asymmetry and local irreversibility.
A single sentence captures the split: the frequency effect generalized, the waveform-local effect did not.
This kind of cross-species mismatch often reveals as much as a perfect replication. It suggests that waveform morphology may be especially sensitive to details of human intracranial sampling or to local circuit organization that is not captured in the same way by subdural recordings in macaques. The result also keeps the interpretation cautious. A waveform feature that is powerful in one setting may not transport cleanly to another without considering anatomy, electrode geometry, and the spatial scale of the measurement.
Why waveform morphology could matter for stimulation timing
The preprint closes with a clinically relevant proposal: waveform morphology may help guide stimulation timing. That idea follows naturally from the mechanistic findings. If synaptic kinetics alter waveform shape and phase-dependent responsiveness, then stimulation delivered at one phase of the cycle may engage a very different circuit state than stimulation delivered a fraction of a cycle later. In disorders where neuromodulation depends on precise timing — including invasive stimulation paradigms that target oscillatory networks — waveform-informed timing could improve how clinicians think about when to intervene, not just where or how strongly.
The work does not claim to have validated a therapeutic algorithm. It instead offers a mechanistic framework. Local cortical physiology shapes the form of the oscillation; the form of the oscillation carries information about local and network interactions; and those interactions may be more phase-sensitive than a frequency-only view would suggest. For future stimulation studies, that chain of reasoning creates a testable hypothesis: timing interventions to the waveform itself may be more informative than using frequency band labels alone.
Because this is a preprint, the findings remain unreviewed and the field should treat the conclusions as provisional. The sample size for the human intracranial recordings was 23 patients, and the abstract does not provide deeper detail on patient characteristics, electrode coverage, or disorder-specific context. The macaque comparison strengthens the mechanistic story, but it also shows that not every relationship observed in humans will survive a change in species or recording modality. Even so, the study offers a focused message for systems neuroscience: waveform shape is not decorative noise. It is a measurable expression of local circuit kinetics, and in the right setting it may help explain how cortical regions talk to one another.
- • bioRxiv preprint DOI 10.64898/2026.11.13.766469: https://doi.org/10.64898/2026.11.13.766469