Variance of human cortical synaptic transmission inferred from non-invasive magnetoencephalography of somatosensory evoked population spikes
Abstract
Non-invasive recordings of somatosensory evoked brain responses (SERs) contain high-frequency (∼600 Hz) evoked responses (hfSERs) which can be recorded at the single-trial level using ultralow-noise magnetoencephalography (MEG). hfSERs were shown to represent cortical population spikes and consist of an early (pre-synaptic) and a late (post-synaptic) component. Here, a statistical model of cortical population spiking reproduced single-trial hfSERs evoked by median-nerve stimulation in 5 healthy subjects. hfSERs were best replicated when variability was added only to the late (post-synaptic) model component, explaining 85% of inter-trial variability across subjects. Adding variability also to the early (pre-synaptic) component caused overfitting in most subjects, yet still indicated larger amplitude and latency variability post-synaptically. This proof-of-concept establishes model-based inference from non-invasively recorded population spikes as a means to estimate synaptic transmission variability in human neocortex. The approach opens a route to studying synaptic function in health and disease where invasive recordings are unavailable.
Related
- Altered Somatosensory Cortical Entrainment in Cerebral Palsy
- Transient cortical activity modulates somatosensory gating during postural control
- Extreme climatic events shape adaptive potential by altering genetic variance for adult survival in a wild population
- Neural oscillations and network specialisation in early childhood captured with next-generation magnetoencephalography
- Behavioral arousal organizes visual cortical population micro-states across the sleep-wake cycle
BibTeX
@article{radzinski2026variance,
title = {{Variance of human cortical synaptic transmission inferred from non-invasive magnetoencephalography of somatosensory evoked population spikes}},
author = {Lukasz Radzinski and Rainer Körber and Jim Barnes and Soudabeh Arsalani and Benjamin Blankertz and Gabriel Curio and Gunnar Waterstraat},
journal = {iScience},
year = {2026},
month = oct,
doi = {10.1016/j.isci.2026.117880},
url = {https://www.cell.com/iscience/fulltext/S2589-0042(26)03263-3},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Variance of human cortical synaptic transmission inferred from non-invasive magnetoencephalography of somatosensory evoked population spikes
AU - Lukasz Radzinski
AU - Rainer Körber
AU - Jim Barnes
AU - Soudabeh Arsalani
AU - Benjamin Blankertz
AU - Gabriel Curio
AU - Gunnar Waterstraat
PY - 2026
DA - 2026/10/20
JO - iScience
DO - 10.1016/j.isci.2026.117880
UR - https://www.cell.com/iscience/fulltext/S2589-0042(26)03263-3
AB - Non-invasive recordings of somatosensory evoked brain responses (SERs) contain high-frequency (∼600 Hz) evoked responses (hfSERs) which can be recorded at the single-trial level using ultralow-noise magnetoencephalography (MEG). hfSERs were shown to represent cortical population spikes and consist of an early (pre-synaptic) and a late (post-synaptic) component. Here, a statistical model of cortical population spiking reproduced single-trial hfSERs evoked by median-nerve stimulation in 5 healthy subjects. hfSERs were best replicated when variability was added only to the late (post-synaptic) model component, explaining 85% of inter-trial variability across subjects. Adding variability also to the early (pre-synaptic) component caused overfitting in most subjects, yet still indicated larger amplitude and latency variability post-synaptically. This proof-of-concept establishes model-based inference from non-invasively recorded population spikes as a means to estimate synaptic transmission variability in human neocortex. The approach opens a route to studying synaptic function in health and disease where invasive recordings are unavailable.
SN - 2589-0042
ER -