Why BOLD Signals Diverge From Neuronal Spiking

You are reading version 1 of this direction, frozen 2026-10-04. The latest version is version 2.

Version 1 of 218 papers33 conceptsfrozen 2026-10-04

In the effort to map the human mind, we often rely on functional Magnetic Resonance Imaging (fMRI) as our primary lens. fMRI measures the Blood-Oxygen-Level-Dependent (BOLD) signal—a proxy for neural activity based on blood flow. For decades, the implicit assumption was simple: more blood flow equals more neurons firing.

However, a growing body of evidence across systems neuroscience and clinical neurology suggests this relationship is far more complex. The metabolic "bill" the brain pays (captured by fMRI) often reflects the cost of synaptic inputs and local processing rather than the actual spiking output of neurons. Understanding this divergence is critical for interpreting how the brain processes everything from optical illusions to motor intent in stroke patients.

The Factory Analogy: Inputs vs. Outputs

To build intuition, imagine a factory. If you measure the factory's activity by looking at its electricity bill, you are seeing the energy consumed by the assembly lines, the lights, and the cooling systems. This is the BOLD signal. However, the electricity bill doesn't tell you exactly how many finished products (spiking outputs) are actually leaving the loading dock. A factory could be buzzing with internal activity—reorganizing parts or preparing for a shift—without shipping a single item.

In neuroscience, this is known as the principle of synaptic-metabolic coupling. As Stelios M. Smirnakis and Nikos K. Logothetis demonstrated in Motion Processing in the Macaque: Revisited with Functional Magnetic Resonance Imaging, the BOLD signal primarily reflects local synaptic processing and sub-threshold integrative activity. This explains why fMRI often reveals higher levels of stimulus selectivity than the spiking rates of individual neurons would suggest: the imaging is picking up the "chatter" of inputs that never quite trigger a full neural discharge.

When Metabolic Signals See What Spikes Miss

This decoupling leads to fascinating discrepancies in how we perceive the world. Consider the phenomenon of "filling in." When we see an illusory motion, our primary visual cortex (V1) shows activity along the path of the illusion, even though there is no physical object there. Lars Muckli and colleagues showed in Primary Visual Cortex Activity along the Apparent-Motion Trace Reflects Illusory Perception that this V1 activity likely represents top-down feedback—synaptic inputs from higher-order brain areas—that fMRI can detect even if the local neurons aren't actually firing spikes.

Similarly, in the study of depth perception, Akiyuki Anzai and Gregory C DeAngelis found that while fMRI suggested certain visual areas were specialized for complex depth computations, direct electrophysiology told a different story. The high metabolic activation likely reflected sub-threshold synaptic inputs or feedback signals rather than the functional output of those neurons (Coding of Stereoscopic Depth Information in Visual Areas V3 and V3A).

The Illusion of Adaptation

One of the most common tools in fMRI research is "repetition suppression"—the idea that if you show a stimulus twice, the brain's response drops, indicating neural adaptation. But does the output of the neurons actually drop?

Joern K. Pomper and Peter Thier investigated this by recording from mirror neurons in monkeys. In Mirror neurons in monkey area F5 do not adapt to the observation of repeated actions, they discovered a startling dissociation: while the local field potentials (the inputs) showed suppression, the actual spiking output of the neurons remained constant. This suggests that many "adaptation" effects seen in fMRI might just be the brain's input machinery becoming more efficient, while the final message sent to the rest of the brain remains unchanged.

Clinical Consequences: Neurovascular Uncoupling

The divergence between blood and neurons isn't just a theoretical problem; it's a life-or-death challenge in neurosurgery. In healthy brains, the relationship between neural firing and blood flow is tight. However, in the presence of high-grade tumors or strokes, this relationship breaks down—a phenomenon called neurovascular uncoupling.

As Nico Sollmann and colleagues explain in Modern preoperative imaging and functional mapping in patients with intracranial glioma, tumors can disrupt local blood flow regulation so severely that the BOLD signal becomes a "false negative." A surgeon might see no fMRI activity in a region and assume it's safe to operate, when in reality, the neurons are firing normally—the blood just isn't responding to them. This same challenge appears in developing Brain-Computer Interfaces for stroke patients, where HEFMI-ICH researchers are now using hybrid EEG-fNIRS systems to capture both the electrical and vascular signals simultaneously to bypass this uncoupling.

A Multi-Scaled View of the Brain

If fMRI is a "spatially low-passed" version of neural activity, as argued by Elias Issa and Alex M. Papanastassiou in Large-Scale, High-Resolution Neurophysiological Maps Underlying fMRI of Macaque Temporal Lobe, then the future of the field lies in multimodal integration.

By combining the high temporal resolution of EEG/MEG with the spatial reach of fMRI, researchers like Christoph Bledowski and David E.J. Linden are beginning to localize the generators of specific brain waves (Localizing P300 Generators in Visual Target and Distractor Processing). The goal is no longer just to see where the brain is active, but to distinguish the "electricity bill" of synaptic integration from the "shipped product" of neuronal communication.

Papers in this version

History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 17 papers
  2. Version 1 · 2026-10-04 · 18 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_metabolic_electrophysiological_decouplin_aec9e810,
  title = {metabolic-electrophysiological decoupling},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/metabolic-electrophysiological-decoupling/rdr_aec9e81035e643f1b5b08c1ffd8984fe},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_why_bold_signals_diverge_from_neuronal_s_aec9e810_v1,
  title = {Why BOLD Signals Diverge From Neuronal Spiking},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/why-bold-signals-diverge-from-neuronal-spiking/rdr_aec9e81035e643f1b5b08c1ffd8984fe/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}