More than two decades after Spencer’s original statement, a special collection published in the journal Frontiers in Network Physiology gathers seventeen contributions that amount, in aggregate, to a partial answer to those debates. The editors of the collection, Hitten P. Zaveri of Yale University, Klaus Lehnertz of the University of Bonn, and Dennis D. Spencer of Yale, note that the clearest progress visible across the papers lies not on the theory’s founding claim, which the contributions do not dispute, but on the practical machinery of finding, characterizing, and treating the network in an individual patient. That shift, from abstract debate to measurable, patient-specific tools, is quietly transforming how epilepsy surgeons and neurologists think about the brain.

The largest cluster of papers addresses the hunt for biomarkers of the epileptic network, and it is where the collection’s empirical weight sits. One review reframes the high-frequency oscillation, a brief electrical event long used to mark epileptogenic tissue, by reading the evidence as showing that these oscillations are not fixed to a single site but propagate along sequential pathways from an onset contact to spread contacts across cortex within milliseconds. On this reading, a high-frequency oscillation is a discrete expression of the network it traverses rather than a local event, and the full propagation zone predicts the resection zone more reliably than any single high-rate contact. That reframing matters because surgeons have long treated individual oscillation hotspots as targets; the new view suggests the pathway matters more than the pinprick.

Building on this premise, one study quantified how consistently each electrode channel’s high-frequency activity synchronizes with the rest of the network using a channel-level connectivity dispersion metric. Tested on 89 and 15 patients from independent centers, the seizure-onset-zone channels showed lower, not higher, dispersion than the surrounding tissue, and the size of that difference tracked surgical outcome. Another team proposed infraslow envelope correlations, measured below 0.2 hertz, as a candidate process that forms the connections of the network. Estimated from three-minute recording windows, these correlations persist across time lags of up to about a minute, a far slower regime than the millisecond-to-second range most connectivity measures operate on, and they rose modestly after medication taper, consistent with heightened seizure vulnerability.

Other biomarker work reaches further toward prediction. One group compared five interictal EEG features for tracking the circadian and multidien cycles that modulate seizure risk, finding that network-derived measures phase-lock to seizures more tightly than the rate of interictal epileptiform discharges and remain more robust when data are subsampled, a property with direct implications for implantable seizure-forecasting devices. Another team developed patient-specific dynamical network models from interictal intracranial EEG to simulate the brain’s response to stimulation in silico, discriminating epileptogenic zone from non-epileptogenic tissue with an area under the curve of 0.95 across 69 patients from five centers. Strikingly, non-epileptogenic channels also showed elevated excitability in patients with successful surgical outcomes, a pattern absent in failures. A related analysis showed that the steady-state behavior of a previously published source-sink index can be derived as the leading eigenvector of a dynamical network model’s state-transition matrix, predicting surgical success with roughly 72 percent accuracy from only five minutes of interictal recording across 65 patients at six centers.

A preictal directed-connectivity analysis applied to a multicenter intracranial EEG dataset of 243 seizures from 61 patients discriminated the seizure onset zone with a higher area under the curve than connectivity measured during the seizure itself, but only once patients were regrouped by whether their dominant preictal pattern was inflow toward, or outflow from, the seizure onset zone. Read as a set, these seven papers converge less on a single biomarker than on a common signature: the seizure onset zone is consistently more legible through its relationship to the surrounding network than through any property intrinsic to the tissue itself. That is evidence for the network account of epileptogenicity that Spencer’s original theory proposed only in outline, now rendered in quantitative detail.

Two papers push toward mechanism at a level the biomarker studies cannot reach directly, and both speak to the question of whether the epileptic network is pathological or physiological. One review argues for the latter reading, showing that the specific ensembles of single neurons active at seizure onset are selectively reactivated during post-seizure slow-wave sleep but not wakefulness, a pattern the author terms seizure-related consolidation, borrowed directly from the mechanisms of ordinary memory formation. On this account, epileptogenesis co-opts a physiological process rather than representing a wholly separate pathology. A compatible conclusion emerges from an in vitro microelectrode-array model of neuron-glia cultures, in which chronic exposure to 4-aminopyridine, a potassium-channel blocker, produces functional connectivity that grows not merely stronger but bimodal in distribution within two days in otherwise-naive culture. Ordinary circuitry, in other words, is sufficient substrate for network pathology to emerge under sustained hyperexcitability.

Several papers document that the practical consequences of network involvement extend well past the tissue a surgeon would resect. In two patients with mesial temporal lobe epilepsy, psychosis emerged during continuous high-frequency anterior thalamic stimulation even as the stimulation suppressed seizures substantially from baseline, suggesting a network disrupted enough to stop seizures can still misbehave in the memory, mood, and sleep domains the same circuitry ordinarily supports. Switching to a duty-cycle stimulation pattern resolved the psychiatric symptoms without sacrificing seizure control, suggesting the two goals are separable in practice even if seizures and psychosis arise from the same network. Using stereo EEG in 48 surgical patients, another study showed that naming, fluency, and recognition-memory deficits track the extent and direction of the interictal irritative network rather than the seizure onset zone, meaning a network’s cognitive footprint can be substantially larger than its seizure-generating footprint. Yet there is also an optimistic counterpoint: in two patients whose scalp EEG showed features of a generalized epileptic encephalopathy, focal resection of a single driving node eliminated every seizure type and the generalized-appearing EEG features resolved completely, suggesting that removing the right node can collapse far more of the network’s pathological behavior than its apparent size would predict.

Two further contributions look past biomarker development toward what the field will need to act on it. One review of current surgical and neuromodulatory strategies through the lens of network theory delivers a sobering observation: more than half of patients continue to experience refractory seizures within a year of surgery, and closing this gap will require shifting surgical targeting from focus removal toward the removal of epileptic-network constituents. Another addresses a different infrastructure gap with the Yale Brain Atlas, a 690-parcel, neuroanatomical landmark-based open-source atlas integrating white-matter connectomes and cortical thickness data from Human Connectome Project subjects with resting-state and meta-analytic functional data, explicitly positioned to standardize neuroanatomy, aid intracranial electrode localization, support multimodal data fusion, and inform decision making for epilepsy surgery.

The collection closes with two contributions that engage the theory at the level of its foundational claims. One survey of the conceptual architecture of the time-evolving epileptic network across timescales from seconds to years reports a finding that complicates any simple account of what the network is: the vertices that turn out to be most central during a seizure are often not in the clinically defined seizure onset zone but in tissue considered unaffected by the pathology, suggesting the field has sometimes been defining, and attempting to control, the wrong part of the network entirely. The final paper proposes, from background intracranial EEG studies, that the network reduces in the main to two nodes, the seizure onset area and the thalamus, with a low-frequency thalamocortical signature governing seizure initiation and a higher-frequency cortico-cortical signature governing spread. A companion analysis of local connectivity supplies empirical support in the form of a distance-graded connectivity profile that peaks not at the seizure onset area itself but in a rim one to five centimeters away, separating surgical outcomes with high accuracy. Whether this parsimonious two-node account can be squared with the more distributed picture, or with the network’s reach into cognition and psychiatry documented elsewhere in the collection, remains substantially unresolved. What is clear is that the real contribution of two decades of network thinking lies less in resolving whether epilepsy is a network disease, a premise these contributions share, than in the practical, patient-specific machinery for locating that network, and in the growing evidence that surgical failure looks less like mislocalization than like incomplete treatment of a network whose true extent these methods are increasingly able to estimate. The task ahead is to close the gap between algorithms that can estimate a network’s extent and a surgical and neuromodulatory practice that has yet to be redesigned around that estimate.

Subject of Research: The network theory of epilepsy and its clinical applications twenty years after its proposal

Article Title: Editorial: The network theory of epilepsy at twenty

Article References: Zaveri, H. P., Lehnertz, K., & Spencer, D. D. (2026). Editorial: The network theory of epilepsy at twenty. Frontiers in Network Physiology, 6, Article 1973521. https://doi.org/10.3389/fnetp.2026.1973521

Image Credits: AI Generated

DOI: 10.3389/fnetp.2026.1973521

Keywords: epilepsy, network theory, seizures, EEG, biomarkers, seizure onset zone, neuromodulation, epilepsy surgery, brain networks, high-frequency oscillations, thalamocortical connectivity, Frontiers in Network Physiology