Summary: A new study reveals that resting-state functional magnetic resonance imaging (rsfMRI), which maps coordinated blood-flow patterns across the brain, can miss certain neural signals that are “invisible” to conventional electrophysiological measurements. By recording rsfMRI and direct neural activity simultaneously in mice, researchers discovered spatial agreement but temporal mismatch between blood-flow-based networks and local neural signals, indicating that rsfMRI may reflect components of brain activity not captured by standard electrophysiology.
These results, which are likely relevant to human studies, refine our understanding of resting-state brain networks (RSNs) and suggest directions for improving how we interpret brain imaging data.
Key Facts:
- rsfMRI infers brain activity from blood flow changes but may omit neural components that are electrophysiology-invisible.
- Simultaneous neural recordings and rsfMRI showed strong spatial correspondence but weak temporal alignment.
- The findings point to an additional component in rsfMRI signals, with implications for human brain network studies and clinical applications.
Source: Penn State
Resting-state functional magnetic resonance imaging (rsfMRI) is widely used to map how different brain regions coordinate at rest by tracking spontaneous blood flow fluctuations. These patterns form resting-state brain networks (RSNs) that are applied across basic and clinical research. However, rsfMRI measures hemodynamic signals rather than direct neuronal firing, leaving an important question unanswered: how do these blood flow patterns relate to the underlying electrical activity of neurons?
A team led by Nanyin Zhang, the Dorothy Foehr Huck and J. Lloyd Huck Chair in Brain Imaging and professor of biomedical engineering at Penn State, investigated this question by recording whole-brain rsfMRI simultaneously with local electrophysiology in rodents. Their study, published in eLife, aimed to directly compare the spatial and temporal characteristics of blood-flow-derived signals and local field potentials (LFPs) recorded from the same brain sites.
Penn State News spoke with Zhang, who also holds appointments in electrical engineering, engineering science and mechanics, and the Huck Institutes of the Life Sciences, about the study’s goals and implications.
How does rsfMRI work and what are its limits?
Zhang: rsfMRI reveals functional connectivity by tracking spontaneous hemodynamic fluctuations that occur when the brain is at rest. The resulting RSNs help identify which areas tend to activate together. But because rsfMRI measures blood flow rather than electrical activity, the exact link between the hemodynamic signal and underlying neuronal events remains unclear, creating a major gap in how we interpret functional brain networks.
How did the study address rsfMRI’s limitations?
Zhang: We performed simultaneous recordings of rsfMRI and electrophysiology from two distinct brain regions in rodents. This enabled us to capture the rsfMRI signal and direct neural activity from the same locations at the same time, allowing a precise comparison of spatial maps and temporal dynamics.
What were the main findings?
Zhang: Spatially, band-specific LFP power maps closely matched the RSNs derived from rsfMRI and could account for up to 90% of spatial variability in those networks. Temporally, however, LFP band-power time series explained only up to about 35% of the local rsfMRI signal’s variance. Removing the LFP-derived time courses from rsfMRI signals had minimal effect on the RSN spatial patterns. In short, while electrophysiological signals can reproduce the spatial structure of RSNs, they do not fully explain the temporal fluctuations observed in rsfMRI.
This discrepancy suggests the presence of an rsfMRI component that is not directly tied to conventional electrophysiological measures — an “electrophysiology-invisible” contribution that plays a substantial role in shaping the rsfMRI signal.
What are the implications for brain imaging and research?
Zhang: The results challenge the assumption that rsfMRI signals are fully explained by electrophysiological activity. If RSNs are influenced by electrophysiology-invisible processes, then interpreting rsfMRI outcomes — for both basic science and clinical applications — requires caution. Further studies are needed to identify these invisible components and to refine models that link neural activity, vascular responses, and the rsfMRI signal.
How do animal rsfMRI studies translate to humans?
Zhang: The basic neural and hemodynamic mechanisms underlying rsfMRI are likely conserved across mammals. Therefore, the findings from rodents are expected to have translational relevance to human rsfMRI research, informing how we analyze and interpret human brain network data.
Funding: This work was supported by the National Institute of Neurological Disorders and Stroke and the National Institute of Mental Health.
About this neuroimaging and brain mapping research news
Author: Sarah Small
Source: Penn State
Contact: Sarah Small – Penn State
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Disparity in temporal and spatial relationships between resting-state electrophysiological and fMRI signals” by Nanyin Zhang et al., eLife.
Abstract
Disparity in temporal and spatial relationships between resting-state electrophysiological and fMRI signals
Resting-state brain networks (RSNs) are widely used to study health and disease, but how these networks map onto underlying neural activity remains uncertain. To probe this issue, the authors conducted simultaneous whole-brain rsfMRI and local electrophysiological recordings in two brain regions of rats.
Spatial analyses showed that band-specific LFP power maps could explain up to 90% of the spatial variability in rsfMRI-derived RSNs. In contrast, LFP band-power time series accounted for at most 35% of temporal variance in local rsfMRI signals, and removing LFP time courses had little effect on RSN spatial patterns.
This spatial–temporal disparity indicates that electrophysiological activity alone cannot fully explain rsfMRI observations and implies the existence of an rsfMRI component contributed by electrophysiology-invisible signals. These findings provide a new perspective on how RSNs should be interpreted.