Sign in to Peak Brain Path
Sign in to access your courses, books, and progress tracker. New here? Signing in creates your account automatically.
Want to explore courses first?
Browse courses and booksSign in to access your courses, books, and progress tracker. New here? Signing in creates your account automatically.
Want to explore courses first?
Browse courses and booksModule 16
Chapter 16 · 1.5 h · 10 quiz items · pass at 80%
This module covers the functional neuroanatomy of regulation under BCIA Domain II as a current-trends adjunct and the IQCB brain-metabolism, glia, neurovascular-coupling, and blood-brain-barrier topics. The quiz proves the learner can define the neurovascular unit, contrast the slow hemodynamic signal with the fast EEG, and say what HEG and fMRI neurofeedback train.
This book has been about the electrical signal, where it comes from and what it means. There is a second signal a brain trainer can train, and it is not electrical at all. It is the flow of blood. Hemoencephalography trains it directly, and functional magnetic resonance imaging neurofeedback trains a close relative of it. To reason about either, and to round out the neuroscience a certification candidate is expected to know, a practitioner needs the physiology of how brain activity and blood flow are coupled. That coupling also brings a cell type this book has so far left in the background, the glia, onto the stage.
The brain is metabolically expensive and has almost no reserves. It cannot store fuel or oxygen the way muscle can, so it depends on a continuous supply delivered by blood, moment to moment, to the regions that need it. The cellular machinery that turns glucose and oxygen into usable energy runs without pause, and the demand it generates is what local blood flow has to track (Magistretti & Pellerin, 1999). When a patch of cortex becomes more active, its demand for oxygen and glucose rises within seconds, and the local blood supply must rise to match. The brain solves this with a tight, local control system that increases blood flow to active tissue, a process called neurovascular coupling.
[[FIG: FIG-24 – The neurovascular unit – HALF PAGE – a capillary, an astrocyte with endfeet on both the vessel and a synapse, and a neuron, labeled HERE]]
Blood flow is not controlled by neurons alone. It is governed by an assembly called the neurovascular unit: neurons, the small blood vessels that feed them, and the glial cells, especially astrocytes, that sit between the two. Astrocytes wrap both synapses and blood vessels, sensing neural activity on one side and signaling the vessels to dilate on the other. The signaling is concrete. When synapses are active, glutamate spills onto the astrocytes wrapping them and raises the calcium concentration inside the astrocyte, which releases vasoactive messengers onto the smooth muscle of nearby arterioles and relaxes it. Active neurons add their own dilating signals, nitric oxide among them. The vessel widens, and flow rises where the work is happening. The detailed account of how this unit develops, operates, and fails in disease has matured into a field of its own (Iadecola, 2017), and the mechanisms by which glia and neurons together control the diameter of brain blood vessels, and therefore local flow, are now well described (Attwell et al., 2010). The practical point for a brain trainer is that glia are not passive packing material. They are active partners in regulating the signal, both the electrical one, by managing the chemical environment of the synapse, and the hemodynamic one, by controlling blood flow.
[[FIG: FIG-25 – Neurovascular coupling versus the EEG – QUARTER PAGE – fast neural activity in milliseconds beside the slow hemodynamic response peaking seconds later, on a shared time axis HERE]]
The same vessels that the neurovascular unit dilates are also a wall. The brain does not let its blood mix freely with its tissue the way most organs do. Between the bloodstream and the neurons sits the blood-brain barrier, and it is built from three of the neurovascular unit's own components. The endothelial cells lining the brain's capillaries are stitched together by tight junctions, protein seals that close the gaps between cells so that substances cannot slip between them. Anything crossing must pass through the cells, by a controlled transporter, rather than around them. Pericytes, contractile cells wrapped around the capillary, sit on the outside of that endothelial tube and help regulate both the barrier and the vessel's diameter. And the astrocyte endfeet, the same endfeet that signal the vessel to dilate, sheathe the capillary almost completely and induce the endothelial cells to maintain their tight seals (Abbott et al., 2010). The barrier is therefore not a separate structure bolted on. It is the neurovascular unit seen from its protective side.
[[FIG: FIG-38 – The blood-brain barrier – HALF PAGE – a brain capillary in cross-section: endothelial cells joined by tight junctions, a pericyte wrapped around the tube, and astrocyte endfeet sheathing the outside; arrows showing glucose and oxygen crossing by transporter while large molecules and immune cells are excluded HERE]]
The barrier admits oxygen, glucose, and a short list of needed molecules through specific transporters and excludes most everything else, including many drugs and most large molecules, which is why brain pharmacology is partly the problem of what can cross. For a brain trainer, two consequences matter. First, the barrier keeps the chemical environment of the synapse stable, which is part of why the resting potential and the rhythms it supports hold steady (Chapter 1). Second, when the barrier is compromised, by inflammation, injury, or vascular disease, immune molecules and plasma proteins leak into the tissue, neuroinflammation follows, and the consequences reach the signal: a brain whose barrier is leaking is a brain whose excitability and metabolic state have shifted, and the EEG shifts with it. The barrier is named in the certification blueprints for this reason, that it is part of how the brain holds the conditions its signaling depends on.
The neurovascular unit introduced the astrocyte as the cell that signals vessels to widen. That is one job among several, and the others are worth naming because they tie the metabolic signal of this chapter back to the electrical signal of the rest of the book. Astrocytes take up the potassium that neurons release when they fire and redistribute it through their network, the spatial buffering that keeps the extracellular environment stable enough for reliable signaling (Chapter 1). They clear glutamate from the synapse and recycle it, converting it to glutamine and returning it to neurons to be remade into transmitter, the glutamate-glutamine cycle that both ends excitatory signaling cleanly and keeps the transmitter supply running (Chapter 15). And they take part in fueling the active synapse: astrocytes take up glucose from the blood, and one model holds that they supply lactate to active neurons as a ready fuel, an astrocyte-neuron metabolic partnership tied to the same activity that drives blood flow (Magistretti & Pellerin, 1999). The common thread is that the astrocyte sits at the junction of the electrical and the metabolic: it manages the ions and transmitters that shape the EEG, and it manages the glucose, lactate, and blood flow that feed the work. The glia are not the background of the signal; they are part of its machinery.
When neural activity rises in a region, neurovascular coupling delivers a surplus of oxygenated blood to it, more than the tissue immediately consumes. The result is a local increase in the ratio of oxygenated to deoxygenated hemoglobin. This is the basis of the blood-oxygen-level-dependent signal that fMRI measures, and of the optical signals that hemoencephalography measures at the scalp. It is an indirect read on neural activity, one step removed, mediated by the vasculature.
Two properties separate it sharply from the EEG. First, it is slow. Where the EEG follows neural events in milliseconds, the hemodynamic response unfolds over seconds, peaking several seconds after the activity that drove it. The shape is stereotyped: flow begins to rise a second or two after the activity, peaks around five to six seconds later, and then dips below baseline before recovering, a post-stimulus undershoot. That whole curve is the hemodynamic response function, and it is not fixed across people. It slows and weakens with age (West et al., 2019), one more reason a hemodynamic measure is read against the person rather than a universal template. Second, it carries different information: the EEG reports the synchrony of electrical activity, while the hemodynamic signal reports metabolic demand and the blood-flow response to it. The two are coupled but not interchangeable, which is why combining them, as in simultaneous EEG and fMRI, reveals more than either alone.
The hemodynamic response described above is evoked: activity rises, and flow follows to meet it. But the brain's small vessels are not still while they wait. They oscillate on their own, slowly and rhythmically, in a phenomenon called vasomotion. The arterioles widen and narrow at roughly a tenth of a hertz, well below one cycle a second, driven by the intrinsic dynamics of vascular smooth muscle and the endothelium and shaped by autonomic and neural input. These are not random flutters. In the cortex the oscillations are organized, often traveling as slow waves along the branches of the vascular tree, and because blood volume and oxygenation are what the hemodynamic signal reads, vasomotion is a large part of what sets the slow signal's baseline. The low-frequency fluctuations that resting-state fMRI and fNIRS measure are, in significant part, this vascular rhythm, entrained to and interacting with neural activity (Mateo et al., 2017).
Two points follow for a brain trainer. First, the slow hemodynamic signal a practitioner trains is riding on a baseline that oscillates on its own, so a hemodynamic recording is never a flat line waiting for activity. It is a rhythm being modulated. Second, the vascular rhythm is plastic. Repeated, spaced engagement can shift its regularity and coherence, and the working hypothesis behind the slow blood-flow modalities below is that they train not just more flow but a more coherent, more flexible vascular oscillation in the engaged region. The detail of how that is turned into a protocol belongs to the coaching literature. The physiology to carry is that the vasculature has a rhythm of its own, near the same tenth-of-a-hertz band as the baroreflex resonance of Chapter 9, and that rhythm is part of what the hemodynamic signal reports.
"Hemoencephalography" is not one device, and the differences are physiological, not just commercial. Four measurement methods read the hemodynamic signal, and they divide by what physical quantity they sense and how deep they reach.
Near-infrared HEG (nIR HEG). Developed in the Toomim lineage, it shines near-infrared light into the forehead and reads how much returns. Because oxygenated and deoxygenated hemoglobin absorb near-infrared light differently, the returning light reports the oxygenation ratio of the blood in the prefrontal cortex beneath the sensor (Toomim et al., 2005). It measures blood oxygenation, optically.
Passive infrared HEG (pIR HEG). Developed in the Carmen lineage, it senses not reflected light but emitted heat: the long-wave infrared the forehead radiates, which tracks the warmth that local metabolism and blood flow produce. It is a thermal proxy for local activity rather than an optical oxygenation read, and it is the modality with the most reported use in migraine work (Carmen, 2005). It measures heat, passively.
Functional near-infrared spectroscopy (fNIRS). The research-grade cousin of nIR HEG. It uses the same near-infrared optical principle but with multiple source- detector pairs and quantitative modeling to estimate concentration changes in oxygenated and deoxygenated hemoglobin across a region of cortex. It is nIR HEG's scientific instrument: same physics, more channels, more rigor.
Functional MRI (the BOLD signal). Not optical or thermal at all. Inside an imaging magnet, fMRI reads the blood-oxygen-level-dependent signal from the whole brain, including the deep structures the surface methods cannot reach. It is the reference hemodynamic measure and the one confined to research and specialized settings.
[[FIG: FIG-37 – Four windows on the hemodynamic signal – HALF PAGE – a comparison grid of nIR HEG (near-infrared optical, oxygenation, prefrontal surface), pIR HEG (passive infrared, emitted heat, prefrontal surface), fNIRS (near-infrared optical, oxy/deoxy concentration, cortical surface, multi-channel), and fMRI/BOLD (magnet, whole brain incl. deep); columns for physical quantity measured and depth reached HERE]]
The dividing lines are physical. The three surface methods (nIR HEG, pIR HEG, fNIRS) reach only the outer cortex, mostly the prefrontal surface under the forehead, because light and heat do not pass deep through the head. fMRI reaches the whole brain because it does not depend on surface optics. And the surface methods split again by what they sense: optical methods read oxygenation, the thermal method reads emitted heat. Which device a practice runs, and how it is configured into a protocol, is the work of the coaching literature. What belongs here is the physiology of what each one is actually measuring.
Because the electrical and hemodynamic signals carry different information on different timescales, the strongest picture comes from reading them together. The EEG offers millisecond timing but poor localization, smeared by volume conduction. The hemodynamic signal (fMRI) offers good localization, including deep structures the EEG cannot reach, but poor timing, blurred by the slow vascular response. Recorded simultaneously, as in combined EEG-fMRI, they complement each other: the EEG says when, the hemodynamic signal says where. This is why research that wants both has moved toward simultaneous recording, and why claims from one method do not transfer cleanly to the other.
It also clarifies what each neurofeedback modality can and cannot reach. Scalp EEG and hemoencephalography both work at the surface, so both are blind to deep structures. Functional-MRI neurofeedback can target a deep region such as the amygdala precisely because it reads the hemodynamic signal inside an imaging magnet, which is also why it is confined to research and specialized settings rather than the clinic. A brain trainer working at the scalp should read the fMRI-neurofeedback literature with that translation in mind: the principle (a brain can learn to regulate a signal fed back to it) carries over, but the reach and the timescale do not.
The vasculature does one more job that belongs here, because it runs on the same vessels, the same astrocytes, and the same slow rhythm. The brain has no classical lymphatic system to drain metabolic waste. Instead it uses a fluid-exchange system, the glymphatic system, that moves cerebrospinal fluid along channels formed by astrocytic endfeet pressed against the blood vessels, the same endfeet that signal the vessel to dilate and that help seal the blood-brain barrier. Fluid is driven inward along the arteries and exchanges with the interstitial fluid around the cells, supplemented by the water channel aquaporin-4 in the astrocyte membrane, and the pulsation and slow oscillation of the vessels, the vasomotion described above, is part of what pumps the exchange along. Clearance, in other words, rides on the vascular rhythm.
The system runs hardest during sleep. In slow-wave sleep the interstitial spaces between cells widen substantially relative to the waking brain, opening more room for fluid to move, and the clearance of metabolic by-products, including the adenosine that builds sleep pressure (Chapter 15) and potentially amyloid, runs markedly faster asleep than awake (Xie et al., 2013). This closes a loop the book has touched from several sides: the slow oscillation of deep sleep (Chapter 7), the adenosine that drives the drive to sleep, and now the clearance that sleep performs all belong to one nightly process. For a brain trainer the physiology explains why sleep architecture matters beyond the subjective experience of rest: a brain with fragmented slow-wave sleep is a brain clearing less and carrying more, and its daytime signal, slowed and heavy, reflects it. What to do about it belongs to the coaching and consumer literature. That the clearance is real, vascular, and sleep-driven belongs here.
Brain fog is a subjective complaint, not a diagnosis, and the physiology in this chapter is what gives it a body. The striking thing is how similar it looks on a recording regardless of cause, whether it follows a concussion, a viral illness, poor sleep, or chronic stress: a brain running in a low-energy state. The contributors are several and they overlap. Metabolic supply can be the problem, mitochondria producing less of the energy the ion pumps of Chapter 1 depend on. Neuroinflammation can be the problem, immune signaling altering synaptic function. Reduced perfusion can be the problem, the neurovascular delivery of this chapter falling short of demand. And impaired clearance can be the problem, the fragmented slow-wave sleep of the section above leaving waste and adenosine elevated. These are not rival explanations so much as a set of ways the same low-energy state arrives.
Read the EEG and the convergence shows. The peak alpha frequency, the speed of the thalamocortical clock of Chapter 6, tends to slow, and a slower clock travels with slower processing. Frontal beta often rises, not because the region is doing more but because it is working harder to hold ordinary function, a compensation that costs energy and produces the paradox of feeling both foggy and wired. And slow waves, delta and theta that belong to sleep, intrude into the waking record, the sign of a cortex trying to rest while awake. None of these is specific to fog by itself. Together they are the signature of a brain low on energy. What to do about it, the sleep, metabolic, and training interventions, belongs to the consumer and coaching literature. The physiology of why the signal looks the way it does belongs here.
A practitioner adds hemoencephalography for a client's brain fog and expects it to behave like the EEG. The physiology resets that expectation. Because the hemodynamic response lags neural activity by seconds, the signal is inherently slower and smoother than an electrical one, and it reports metabolic demand rather than electrical synchrony, exactly what a low-energy, underperfused state would fall short on. Knowing it is a slow, metabolic signal rather than a fast, electrical one is what keeps expectations correctly calibrated. How the modality is run belongs to the coaching literature.
What this means for the signal: the EEG is the brain's electrical signal, but the same active tissue also drives a slower, metabolic signal carried by blood flow and governed by the neurovascular unit, glia included. A brain trainer who works with hemoencephalography or reads the fMRI-neurofeedback literature is working with this second signal, and the two together give a fuller picture of the brain at work than either does alone.
Key points
In one sentence: the same active tissue drives a second, slower signal carried by blood flow, with its own rhythm, its own barrier, and its own nightly clearance.
Check yourself
Ch 2 (electrical signal for contrast), Ch 5 (glia introduced), Ch 15 (astrocytes and the synapse), Coaching (HEG modality), Field Guide (EEG vs other neuroimaging modalities).
Added 2026-05-29 to close a breadth gap: the book trains EEG providers, but HEG and
fMRI neurofeedback train the hemodynamic signal, and brain metabolism / glia /
neurovascular coupling are part of comprehensive neuroscience for both boards. See
meta/strategy-drivers-from-analysis.md.