The brain’s blood supply is closely linked to neural activity, a relationship that underpins widely used brain‑imaging techniques such as fMRI. Yet for decades, neuroscientists have debated whether this link—known as neurovascular coupling—varies across brain regions and brain states, or whether it follows a common rule. A new study from researchers at UCL Institute of Ophthalmology and Queen Square Institute of Neurology and published in Nature now offers a unifying answer.
By simultaneously measuring brainwide neural activity and blood volume in mice, researchers have discovered that brainwide blood flow is driven by the combined activity of two opposing populations of neurons. These populations, which are distributed throughout every brain region, have opposing relations to brain state and distinct relationships to blood supply. Together, they account for widespread fluctuations in blood volume across the brain.
Previous studies of neurovascular coupling often relied on bulk measures of neural activity and focused on different brain regions or behavioural events, making results difficult to compare. Some reports suggested that increases in neural activity always lead to stronger local blood flow, while others found weaker, or even negative, relationships in certain regions or states, such as sleep or rest. These inconsistencies raised the possibility that blood supply might be shaped by factors such as arousal or attention independently of neuronal firing.
To overcome these limitations, the researchers took a brainwide approach. Using functional ultrasound imaging (fUSI), they measured changes in blood volume across large swathes of the mouse brain. In parallel, they recorded the activity of thousands of individual neurons using high‑density Neuropixels probes.
The imaging revealed that arousal, tracked using whisking behaviour as a reliable marker, was associated with stereotyped, brainwide changes in blood volume. Crucially, the neural recordings showed that this global vascular signal could be explained by two distinct groups of neurons: one population increased its activity during arousal, while the other decreased it.
These two populations exhibited different haemodynamic response functions (HRFs), which describe how changes in neural activity translate into changes in blood supply. Both HRFs were positive, meaning increases in firing rate were linked to increases in blood volume, but with different dynamics. When the contributions of both populations were combined, they accurately predicted blood volume fluctuations across different brain states.
Further brainwide recordings revealed that both populations coexist in every brain region examined, from cortex to midbrain. Apparent regional differences in blood flow dynamics, the researchers found, largely reflected differences in the balance between these two neural populations rather than fundamental regional differences in neurovascular coupling.
Professor Matteo Carandini (UCL Institute of Ophthalmology), a co-author on the study, said: “Our findings reveal that the relation between brain activity and blood flow in the brain is written in stone: it is the same in all brain regions and in all brain states, from sleep to high alertness. This is great news, as it gives us new confidence in trusting fMRI scans, which measure blood flow to estimate brain activity.
“However, we also discovered that blood flow depends differently on the activity of two types of neurons. This makes it a bit harder to interpret fMRI scans in terms of brain activity: if a patient shows more blood flow in a certain brain region than a healthy participant, does it mean that it is one type of neuron that is more active? Or is it the other type of neuron? This is the next frontier that the field needs to conquer. The problem already existed; at least now we know about it.”
Lead author Dr Agnès Landemard (UCL Institute of Ophthalmology) said: “Our results suggest that what we measure with brain‑imaging techniques like fMRI reflects the combined activity of two different neural populations, not a single uniform signal. That helps explain why neurovascular coupling has seemed so inconsistent across studies.”