{
  "id": 6329387,
  "title": "Tracking propagating cortical activity in MEG/EEG with a bilinear state-space model",
  "url": "https://urgent.news/2026/09/08/tracking-propagating-cortical-activity-in-meg-eeg-with-a-bilinear",
  "topic": "science",
  "section": "Science",
  "published": "2026-09-08T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.09.02.748579v1?rss=1"
  },
  "original_language": "en",
  "account": "Magnetoencephalography (MEG) and electroencephalography (EEG) are powerful tools for studying large-scale brain activity. However, tracking the movement of cortical waves using these non-invasive methods poses a significant challenge. Conventional inverse solutions assume that sources are spatially and temporally separate, limiting their ability to capture the continuous movement of cortical waves and often misinterpret stationary sources as false signals. To overcome this limitation, we introduce a new dynamic state-space framework that explicitly accounts for the inseparability of spatial and temporal aspects of neural activity. Our method models the sensor signal as a bilinear combination of two evolving states: a fast, narrowband stochastic oscillator representing the electrical dynamics, and a slowly changing spatial pattern that moves through a subspace identified through singular value decomposition. These two states are estimated simultaneously using an Unscented Kalman Filter. To assess the performance of our approach, we simulated MEG data and analyzed real-world resting-state EEG recordings focused on the occipital alpha rhythm. The simulations demonstrated that our method accurately captured both the electrical time courses and spatial trajectories of cortical waves, even under conditions of varying signal-to-noise ratios, fast-moving sources, and different cortical geometries. Importantly, the model was able to distinguish between genuine propagation and static, coherent dipoles. Furthermore, when applied to empirical MEG and EEG data of the occipital alpha rhythm, our framework produced anatomically realistic and temporally coherent propagation patterns that explained significantly more variance in the sensor data than traditional static baselines. By incorporating the evolving source geometry directly into the inverse solution, this dynamic state-space framework offers a promising, proof-of-concept tool for the non-invasive study of large-scale brain dynamics.",
  "summary": "Magnetoencephalography (MEG) and electroencephalography (EEG) are ideal for studying macroscopic neural dynamics, but non-invasive tracking of cortical traveling waves remains a major methodological challenge. Traditional inverse solutions assume spatiotemporal separability, restricting sources to fixed spatial topographies. Consequently, they struggle to capture the continuous spatial migration…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}