Bilateral detection, prediction, and lateralization of seizures using white matter

To determine how the corpus callosum handles interhemispheric seizure propagation and to assess its potential as a strategic recording locus for state-dependent biomarkers.

Epileptiform activity was induced in vitro and in vivo using 4-aminopyridine. Propagation dynamics were examined using cross-correlation analysis. Signal features were quantified using root mean square amplitudes, Shannon entropy, and spectral coherence. A previously unreported callosal evoked response, the corpus callosum evoked echo, was characterized pharmacologically (hypocalcemia, hypoxia) and anatomically (callosal disconnection) to determine synaptic dependence and cortical origin. The echo amplitude was evaluated as a predictive biomarker using logistic regression.

Epileptiform discharges propagated through the corpus callosum with delays consistent with polysynaptic cellular transmission (24.81 ± 4.09 ms in vitro), and directionality was resolved in vivo. Although the propagation attenuated the seizure signal power, the spectral content of seizures was largely preserved within the corpus callosum. Following epileptogenic induction (pre-seizure), the corpus callosum signal's root mean square increased significantly, supporting its value as a physiological marker of propagated seizure activity. The corpus callosum-evoked echoes were temporally distinct from compound action potentials, were eliminated by synaptic blockade and callosal disconnection, and were selectively vulnerable to metabolic compromise, supporting their dependence on a cortico-callosal feedback loop. The echo amplitude increased during the pre-ictal period, achieving an area under the curve of 0.77 for seizure prediction in vitro.

These findings identify the corpus callosum as a physiologically active component of epileptic networks rather than a passive transmission pathway. By revealing a previously unrecognized cortico-callosal response and demonstrating its predictive potential, this work supports a network-level framework for epilepsy in which white matter structures contribute directly to seizure generation and may provide targets for physiological monitoring and intervention.

White matter tracts, such as the corpus callosum, have long been dismissed as passive "wires" that merely route seizure activity. This study challenges that paradigm by demonstrating that white matter signals reflect complex epileptic network behavior. By tapping directly into the corpus callosum, we achieved bilateral epilepsy monitoring and a new approach to pinpointing seizure origins within the brain. Furthermore, we uncovered a novel network response called the corpus callosum evoked echo (ccEE), which shows promise as a biomarker for seizure onset. These findings establish white matter as a critical therapeutic target for early seizure forecasting and next-generation, closed-loop neuromodulation.