EEG Biomarkers for Concussion

Three Million Invisible Injuries Per Year

Somewhere right now, a college linebacker is sitting in a neurologist's office staring at an MRI that looks perfectly normal. His brain feels like it's running through wet cement. He can't hold a thought for more than a few seconds. Reading gives him headaches. But the scan says everything is fine.

Here's what nobody told him: MRI photographs the architecture of the brain. It's looking for cracks in the walls, leaks in the pipes, structural collapse. And in a typical concussion, the architecture is fine. The damage is happening at a level MRI was never designed to see. It's happening in the electricity.

Your brain runs on electrical signals. Roughly 86 billion neurons firing in coordinated patterns, producing oscillations that we can measure through the skull using electroencephalography, or EEG. And when a concussion shakes those neurons loose from their normal rhythms, the disruption shows up in the EEG like a bad radio signal. Static where there should be clarity. Silence where there should be activity. Timing thrown off by fractions of a second that cascade into real cognitive problems.

Over the past two decades, researchers have identified a specific set of EEG changes that consistently appear after concussion. These aren't vague, hand-wavy "brain changes." They're quantifiable biomarkers with documented sensitivity, specificity, and clinical utility. And they're quietly reshaping everything from sports medicine protocols to military readiness assessments to disability evaluations.

The research is deep, well-replicated, and frankly underappreciated outside of specialized clinical circles. Let's change that.

A Quick Primer on What EEG Actually Measures

If you already know how EEG works, skip ahead. But if the phrase "theta power" sounds like something from a video game, stay with me for sixty seconds.

Your neurons communicate using tiny electrical pulses. When large populations of neurons fire together in synchronized patterns, they produce oscillations, rhythmic waves of electrical activity, that are strong enough to detect through the skull. An EEG device places electrodes on the scalp and records these oscillations.

The oscillations come in different frequency bands, each associated with different brain states:

• Delta (0.5-4 Hz): Deep sleep. Very slow, very large waves.
• Theta (4-8 Hz): Drowsiness, light sleep, some meditative states. Slow and rolling.
• Alpha (8-13 Hz): Relaxed wakefulness, especially with eyes closed. The brain's "idle hum."
• Beta (13-30 Hz): Active thinking, focus, problem-solving. Quick and busy.
• Gamma (30-100 Hz): High-level information processing, binding of perception. The fastest rhythm.

In a healthy, awake brain performing a cognitive task, you'd expect to see moderate alpha, active beta, low theta, and minimal delta. The balance between these bands tells you a lot about how the brain is functioning. And after a concussion, that balance gets wrecked in very specific ways.

The Biomarker Constellation: What Concussion Does to Brainwaves

Researchers didn't find just one EEG change after concussion. They found a whole pattern, a constellation of interrelated disruptions that, taken together, form something close to an electrical fingerprint of brain injury.

Theta Surges: The Brain Running on Fumes

The most consistent finding across concussion EEG studies is an increase in theta power, particularly over frontal and temporal regions. This isn't a subtle wobble in the data. Quantitative EEG (QEEG) studies have documented theta power increases of 30% to 50% compared to pre-injury baselines.

Why does theta surge? It comes back to what's happening at the cellular level. When the brain sustains a concussive impact, axons (the long communication cables that connect neurons) get stretched by rotational shearing forces. Their membranes become leaky. Ions flood in and out of channels that are supposed to open and close with precision. The neuron's internal machinery goes into overdrive trying to restore balance, burning through glucose and ATP at an unsustainable rate.

The result is an energy crisis. Neurons that are running low on metabolic fuel can't maintain the fast, coordinated firing patterns that produce alpha and beta rhythms. They slow down. They drop into theta-range oscillations. It's the electrical equivalent of a city during a brownout: the lights are still on, but they're flickering, dimming, operating at a fraction of their normal capacity.

A 2017 study by Munia et al. in Clinical Neurophysiology showed that this theta increase appears within hours of injury and can persist for weeks. Crucially, it was detectable even in athletes who reported feeling "back to normal" within a few days.

Alpha Collapse: The Idle Rhythm Goes Quiet

alpha brainwaves are your brain's resting rhythm, strongest over posterior regions when you're relaxed with eyes closed. They reflect healthy functioning of the thalamocortical circuits, the feedback loops connecting the thalamus (the brain's relay station) to the cortex.

After a concussion, alpha power drops. This is one of the most replicated findings in the literature. A meta-analysis in Clinical Neurophysiology (2021) confirmed reduced posterior alpha in over 80% of concussion cases assessed with QEEG.

But the real diagnostic value isn't just in alpha power itself. It's in alpha reactivity: the way alpha responds to eye opening and closing. In a healthy brain, alpha power increases when you close your eyes and decreases when you open them. This toggle is a sign that your arousal systems are functioning properly.

After a concussion, this reactivity dulls. The alpha rhythm becomes sluggish, less responsive, almost flat in severe cases. It's like a thermostat that's stopped regulating temperature. The heater is still there, but it's no longer responding to changes in the room.

Coherence Breakdown: The Network Fractures

Here's where the research gets genuinely fascinating, and where EEG reveals something no other neuroimaging method can show in a clinician's office.

Coherence measures how synchronized the electrical activity is between different brain regions. When your visual cortex and prefrontal cortex need to cooperate on a task (say, reading this sentence and deciding what it means), their oscillations lock into phase with each other. High coherence means strong communication between regions. Low coherence means the connection is degraded.

After a concussion, coherence drops in a pattern that maps directly onto the white matter tracts most vulnerable to rotational injury. The longest axon bundles, particularly frontal-parietal connections and interhemispheric fibers crossing through the corpus callosum, are the most susceptible to shearing forces. And those are exactly the pathways where coherence drops the most.

Here's the finding that should make you sit up: a 2019 study in Brain Injury tracked coherence in college athletes after concussion. Their symptoms resolved in an average of 10 days. Their coherence values didn't normalize for an average of 45 days. The athletes felt fine for over a month while their brains were still measurably disconnected.

That gap between "feeling better" and "being better" is the central problem of concussion management. And coherence measurement might be the best tool we have for closing it.

P300 Delays: The Brain's Processing Speed, Measured in Milliseconds

The P300 is an event-related potential (ERP), a specific brainwave response that fires approximately 300 milliseconds after you encounter something unexpected or relevant. In a standard oddball paradigm, you might see a repeated stimulus (the letter "X") with occasional rare stimuli (the letter "O") mixed in. Each time the rare stimulus appears, your brain generates a large positive voltage deflection around 300 milliseconds later.

This response is a direct readout of three cognitive processes: attention allocation, working memory updating, and processing speed. All three take a hit after concussion.

Two things happen to the P300 in concussed individuals:

Amplitude shrinks. The brain is recruiting fewer neural resources to process the significant event. Think of it as whispering when you should be shouting.

Latency stretches. Instead of peaking at 300 milliseconds, the P300 might peak at 350, 400, or even 450 milliseconds. The brain is taking measurably longer to process the same information.

Broglio and colleagues published a landmark study showing that P300 changes could distinguish concussed athletes from healthy controls with 81% accuracy, outperforming both symptom checklists and standard neuropsychological tests. And like coherence, P300 abnormalities persisted well beyond symptom resolution.

The Complete Biomarker Map

Pulling all of these findings together, here's what the research shows:

Notice something? Almost every biomarker takes longer to normalize than symptoms take to resolve. This isn't a minor discrepancy. It's a fundamental mismatch between subjective experience and objective brain function. And it has life-or-death implications for return-to-play decisions, which we'll get to shortly.

QEEG Discriminant Analysis: Pattern Recognition Meets Brain Injury

Individual biomarkers are useful, but the real leap came when researchers asked: what if we analyzed all of these features simultaneously? What if we fed the entire EEG pattern into a statistical classifier and asked it to distinguish concussed brains from healthy ones?

This is the idea behind QEEG discriminant analysis, and it's produced some of the most impressive accuracy numbers in concussion diagnostics.

How It Works

QEEG discriminant analysis starts with a normative database, a large collection of EEG recordings from healthy individuals, stratified by age and sometimes sex. Each recording is decomposed into dozens or hundreds of features: absolute and relative power in each frequency band at each electrode, coherence between every pair of electrodes, phase relationships, amplitude asymmetries, power ratios, and more.

When a potentially concussed individual's EEG is recorded, the same features are extracted and compared to the normative database. A discriminant function, essentially a weighted combination of the features that best separates concussed from non-concussed recordings, classifies the new recording.

The Thatcher Discriminant Function

Robert Thatcher's work has been among the most cited in this field. His discriminant function, developed using the NeuroGuide normative database, evaluates phase, coherence, and amplitude asymmetry across multiple electrode pairs to produce a classification score. In his published validation studies, the function achieved classification accuracies between 92% and 96% for separating mild traumatic brain injury (mTBI) from healthy controls.

Those numbers deserve context. A standard clinical MRI identifies abnormalities in roughly 28% of symptomatic concussion patients. A QEEG discriminant function identifies abnormalities in over 90%. That's not a modest improvement. It's a fundamentally different level of sensitivity.

Validation and Criticism

It would be irresponsible to present those numbers without mentioning the debate around them. QEEG discriminant analysis has critics, and their concerns are legitimate.

Sample size concerns. Some validation studies used relatively small samples, which can inflate accuracy estimates. Larger, multi-site replication studies are needed to confirm the exact sensitivity and specificity numbers.

Normative database limitations. The accuracy of discriminant analysis depends entirely on the quality and representativeness of the normative database. If the database doesn't adequately represent the patient's demographic group, comparisons lose validity.

Clinical adoption barriers. Many neurologists trained in visual EEG interpretation are skeptical of quantitative approaches. The American Academy of Neurology has been cautious in its endorsement, calling for more standardized validation before widespread clinical use.

Classification vs. diagnosis. A discriminant function can tell you that an EEG pattern is statistically consistent with concussion. It cannot tell you that a person has a concussion. Diagnosis requires integrating EEG findings with clinical history, neurological examination, symptom assessment, and cognitive testing.

That said, the overall direction of the evidence is clear. Multiple independent research groups using different databases and different analytical approaches have converged on the same conclusion: EEG-based quantitative analysis detects concussion-related brain changes with sensitivity that structural imaging cannot match. The debate is about the exact numbers and the path to clinical standardization, not about whether the signal exists.

Brainwave data, captured at 256Hz across 8 channels, processed on-device. The Crown's open SDKs let developers build brain-responsive applications.

Explore the Crown

Return-to-Play: The Highest Stakes Application

If there's one area where EEG concussion biomarkers have the most immediate, tangible impact, it's in return-to-play decisions for athletes.

The current consensus protocol, developed by the Concussion in Sport Group (CISG), uses a graduated return-to-play framework based primarily on symptom resolution. An athlete progresses through stages of increasing physical and cognitive demand. If symptoms don't return at each stage, they advance.

The problem is now obvious: symptoms and brain function don't recover on the same timeline. An athlete can sail through every stage of the symptom-based protocol while their EEG still shows elevated theta, suppressed coherence, and delayed P300 responses. And a brain that hasn't fully recovered is catastrophically vulnerable to repeat injury.

Second Impact Syndrome

When a second concussion occurs before the first has fully healed, the consequences can be severe. Second impact syndrome, though rare, can cause rapid, massive cerebral edema (brain swelling) that is sometimes fatal. Even without such extreme outcomes, repeated sub-threshold impacts on an incompletely recovered brain accelerate the accumulation of damage associated with chronic traumatic encephalopathy (CTE).

This is why the gap between symptomatic recovery and electrophysiological recovery isn't an academic curiosity. It's a safety issue measured in permanent brain damage and, in worst cases, lives.

EEG-Informed Clearance Protocols

Progressive sports medicine programs have begun integrating QEEG into their return-to-play frameworks. The approach typically follows four steps:

1. Pre-season baseline. Every athlete records a standardized EEG session before the season begins. This captures their individual "normal" brainwave signature, including resting power spectra, alpha reactivity, coherence patterns, and (in some protocols) P300 timing.

2. Post-injury comparison. After concussion, the same recording protocol is repeated and compared to the athlete's own baseline. Individual comparison is critical because normal EEG varies substantially between people.

3. Longitudinal tracking. EEG is recorded at regular intervals (weekly or biweekly) throughout recovery. Key metrics are plotted against the baseline trajectory.

4. Dual-criteria clearance. Return to full contact is authorized only when both clinical symptoms have resolved AND EEG biomarkers have returned to within a defined threshold of the individual's baseline.

A 2023 study in the British Journal of Sports Medicine compared outcomes of athletes cleared using EEG-informed protocols versus symptom-only protocols. The EEG group had a 67% lower rate of concussion recurrence. That single statistic may be the most compelling argument for adding objective brain monitoring to concussion management.

Beyond the Playing Field: Where This Research Is Heading

The sports medicine application gets the headlines, but EEG concussion biomarkers are finding traction in several other areas that deserve attention.

Military and Occupational Settings

Blast-related mild TBI is the signature injury of modern warfare. The Department of Defense has invested heavily in EEG-based screening tools that can be deployed in field settings, where MRI is unavailable and symptom reporting is complicated by the fact that soldiers may minimize symptoms to stay with their units. Portable EEG devices with automated discriminant analysis could provide objective assessment within hours of blast exposure.

Post-Concussion Syndrome and Persistent Symptoms

Roughly 10% to 30% of concussion patients develop persistent post-concussion symptoms lasting months or years. These individuals often face skepticism because their structural imaging is normal. QEEG studies have consistently found abnormal brainwave patterns in this population, providing objective evidence of ongoing brain dysfunction. This has implications for disability evaluations, treatment planning, and simply validating the patient's experience.

Pediatric Concussion

Children's brains are still developing, which complicates both the injury and the recovery process. Normative EEG databases for pediatric populations are being refined, and early research suggests that EEG biomarkers may be even more sensitive to concussion in children and adolescents than in adults. Given the higher stakes (developing brains are more vulnerable to repeat injury), objective recovery monitoring in young athletes may become standard practice within the decade.

Subconcussive Impact Research

Perhaps the most provocative line of research involves impacts that don't produce diagnosable concussions at all. Repetitive subconcussive impacts, the routine hits of a football lineman's daily practice, have been shown to produce cumulative EEG changes over a single season. Theta power drifts upward. Coherence degrades slightly. The changes are smaller than those seen in diagnosed concussion but are statistically significant and, in some studies, correlate with subtle cognitive declines on neuropsychological testing.

This research is still early, but if it holds up, it could fundamentally change how we think about contact sports. The question would shift from "has this person had a concussion?" to "is this person's cumulative brain electrical signature drifting away from baseline?"

Tracking Recovery With Consumer EEG

Here's where the research intersects with real-world accessibility. Clinical QEEG assessments typically happen once or twice during recovery. They require a clinic visit, a technician, a full-cap setup with 19 or more electrodes, and specialized software. The result is a snapshot, valuable but limited.

Consumer-grade EEG devices with sufficient channel count and sampling rate can fill in the gaps between clinical assessments, providing the longitudinal data that recovery tracking truly requires. Because concussion recovery isn't linear. There are good days and bad days, and a single snapshot can be misleading. Daily or weekly recordings build a trajectory that tells you far more than any single session can.

What 8-Channel Consumer EEG Can Capture

An 8-channel EEG device sampling at 256Hz has the resolution to track several validated concussion biomarkers:

Power spectrum changes. Theta surges and alpha suppression show up clearly in frequency-band analysis. Tracking daily shifts in theta/alpha ratio gives you a quantitative recovery curve that "how do you feel?" can't match.

Coherence patterns. With electrodes distributed across frontal, central, and parietal regions of both hemispheres, 8-channel EEG captures the most clinically relevant coherence pathways: frontal-parietal and interhemispheric connections.

Alpha reactivity. A simple eyes-open/eyes-closed protocol recorded daily provides one of the most informative single metrics for thalamocortical recovery. No special stimuli needed. Just close your eyes for two minutes, open them for two minutes, and let the EEG do the math.

Trend analysis over weeks. This is where consumer EEG may offer something clinical assessment cannot. Not better spatial resolution or diagnostic authority, but temporal density. Enough data points, recorded consistently over enough days, to reveal the actual trajectory of recovery rather than guessing at it from a couple of snapshots.

The Neurosity Crown sits at electrode positions CP3, C3, F5, PO3, PO4, F6, C4, and CP4, spanning frontal, central, and parietal cortex across both hemispheres. That layout captures the coherence pathways most affected by concussion. Its 256Hz sampling rate cleanly separates theta, alpha, beta, and gamma bands. And the open SDKs (JavaScript and Python) give developers and researchers the tools to build custom monitoring protocols, automate metric extraction, and integrate with analysis pipelines through BrainFlow and Lab Streaming Layer.

The Signal Was Always There

For most of the history of neurology, concussion was a clinical diagnosis made through conversation and observation. A doctor asked you questions. Watched you walk. Tested your reflexes. And then made a judgment call based on training and experience.

That process works, but it has a blind spot the size of a continent. The subjective experience of recovery consistently runs ahead of objective brain recovery. People feel fine while their neural networks are still mending. They return to classrooms, offices, and playing fields while their brainwaves still carry the signature of injury.

The EEG biomarkers we've covered here aren't new discoveries. Researchers have been documenting theta surges and coherence drops after concussion since the 1990s. What's new is the convergence of validation, accessibility, and technology that's finally bringing these findings out of research journals and into practical use.

QEEG discriminant functions can classify concussed brains with accuracy that makes structural imaging look like a coin flip. Return-to-play protocols incorporating EEG data produce measurably better outcomes. And for the first time, the hardware needed to track these biomarkers longitudinally doesn't require a hospital or a research lab. It fits on your head.

The brain's electrical signals have always contained the truth about concussion. We just hadn't built the right tools to listen. Now we have. And the gap between what your symptoms tell you and what your brainwaves reveal is exactly the space where the next generation of concussion management will live.

Your brain has been trying to tell you how it's really doing. The question is whether you're ready to hear it.