Skip to main content
WBF What is
brain fog?
Support WBF Take quiz

Test guide Specialist assessment

Quantitative EEG (qEEG): Brain Maps, Z-Scores, and Results

qEEG turns a scalp EEG into power, frequency, connectivity, ratio, and z-score measurements. There is no universal normal qEEG brain map. Before reading the colors, check the raw EEG, recording quality, montage, state, clean duration, software, and exact reference database. A statistical outlier is not automatically a disease or the cause of brain fog.

21 versus 19 A standard clinical EEG uses all 21 IFCN-recommended electrodes; major research norms often analyze 19 scalp channels. z = 2 Two database standard deviations from its fitted mean, not a 2 percent risk or a diagnosis. 1,564 recordings The multinational HarMNqEEG project spanned 9 countries and 14 devices, while still finding batch effects. Bring with it Raw EEG interpretation, method details, sleep, medicines, symptoms, real-life failures, and the decision the report changes.
01

What can qEEG measure, and what can it not answer?

A qEEG report can look exact because it uses colors, percentiles, and z-scores. First check whether the recording was clean and whether the result changes a real medical decision. Sleepiness, eye or muscle movement, poor electrode contact, medicines, caffeine, migraine, pain, anxiety, and illness can change the map. Age, device, reference, and the comparison database also matter.

Power and frequency

How much measured activity sits in named frequency bands

Absolute and relative delta, theta, alpha, and beta power, plus peak alpha frequency, depend on the state, band boundaries, reference, artifact removal, age, and software.

Relationships

How readings compare across scalp locations

Asymmetry, coherence, phase, and other connectivity measures are mathematical relationships. Scalp location is not a direct picture of one brain area's function or damage.

Database distance

How far a value sits from one reference model

A z-score places a measure in standard-deviation units relative to a named model. Database age, country, device, montage, reference, state, and batch can change that comparison.

A colorful outlier is a question, not a diagnosis

The map cannot by itself diagnose ADHD, concussion, Long COVID, ME/CFS, depression, dementia, epilepsy, dishonesty, or a cause of brain fog. It also cannot choose neurofeedback, medicine, supplements, or a treatment package. The raw EEG, technical quality, history, examination, standard diagnostic workup, and decision the result changes come first.

Save this test

Save the report, test method, and symptoms it was compared with

File the raw EEG conclusion and full quantitative method with the highlighted measures, z-scores, recording conditions, symptoms, function, and clinician explanation. A later reader can then see the method behind each color or number.

My Fog stores the report and details you enter. It does not analyze the raw EEG or calculate z-scores. A clinician must diagnose conditions, check the map, and choose treatment or repeat testing.

02

How do age, sex, pregnancy, sleep, medicines, illness, device, and database change qEEG results?

qEEG values change with age, recording state, device, and test method. Some databases include sex, but there is no single brain-map cutoff for males, females, pregnancy, or any life stage.

Babies and young children

EEG changes quickly during early development. A child needs a pediatric neurophysiology service, a developmentally suitable recording, and a database that actually includes the child's age. Do not apply an adult color map or adult frequency expectations.

Children and teenagers

Children have more low-frequency delta and theta activity than adults, and alpha features change with maturation. The FDA NEBA clearance covers ages 6 through 17 for one narrow ADHD-aid use, not every qEEG map. Pediatric methods still vary enough that age alone cannot rescue a poorly matched database.

Adult women and men

Ask which reference database and sex categories were used, and apply the same technical quality checks to every result. Some measures differ by sex in research datasets, but there are no universal separate male and female diagnostic maps. Symptoms, function, sleep, medicines, migraine, mood, and the clinical question remain more important than a demographic label alone.

Pregnancy and the months after birth

There is no established generic pregnancy or postpartum qEEG diagnostic range. Record gestational week or months after birth, sleep disruption, headache, blood pressure, anemia, thyroid symptoms, medicines, and any seizure history. New severe headache, visual change, weakness, seizure, marked confusion, or very high blood pressure needs maternity or emergency assessment, not a commercial map.

Older adults

EEG features change with age, so an older person needs an age-matched group. Medicines, sleep, hearing, vision, blood-vessel disease, seizure risk, confusion, and brain disease also need review. A map alone does not explain new loss of function.

03

What happens during qEEG brain mapping, and what should you do before the appointment?

Ask which test you're actually getting: a standard clinical EEG with quantitative analysis, a research qEEG, a commercial brain map, or a specific cleared device. Ask which clinical question the result is meant to answer and what decision would change for each possible result.

Confirm who will review the raw EEG. Historical AAN and ACNS guidance says clinical qEEG analysis should only happen alongside the routine EEG analysis. A color map alone can't show whether an apparent finding came from eye movement, muscle, drowsiness, a loose electrode, or real brain activity.

Follow the service's written instructions. Current NHS and Australian patient guidance usually asks for clean, dry hair with no gel, wax, oil, spray, or other hair product. Hair products and tightly fixed hair pieces can make scalp contact harder.

Unless the service gives a different plan, eat and drink normally and take usual medicines. Ask specifically about caffeine, nicotine, cannabis, alcohol, sleep medicines, stimulants, antiseizure medicines, and sedating medicines. Also ask about any medicine change near the recording. Do not stop a prescription or withdraw from a substance on your own.

Do not deprive yourself of sleep unless a clinician has ordered a sleep-deprived EEG and given safety instructions. The NHS and Healthdirect advise not driving after a sleep-deprived EEG. Arrange transport if that is the protocol.

Bring glasses, hearing aids, an up-to-date medicine list, earlier EEG or imaging reports, and a short description of the actual episodes or thinking problems to check. Record recent sleep, illness, fever, pain, migraine, anxiety, food, caffeine, alcohol, cannabis, nicotine, and medicine timing.

Ask about the planned recording state, such as eyes closed, eyes open, resting, drowsy, asleep, or doing a task. Ask how many minutes of clean data the report will analyze. One multinational normative project required at least 1 minute of artifact-free eyes-closed resting EEG, but that research rule is not a universal clinical minimum.

Tell the service about a first seizure, repeat spells, fainting, a head injury, or a severe new headache. Also report pregnancy, implanted or metal devices, scalp wounds, sensory needs, tremor, tics, movement problems, or trouble staying still. These facts may change the test or safety plan.

01

Place the scalp electrodes

The technologist measures the head, prepares the scalp, and attaches electrodes. ACNS minimum technical guidance describes all 21 electrodes and placements recommended by the IFCN for a standard clinical EEG. Research qEEG datasets often analyze 19 scalp channels, which is not the same statement.

02

Record named states

The recording may include eyes closed, eyes open, breathing, flashing lights, drowsiness, sleep, or a task. Keep the exact state and duration with every result, because a value from one state isn't directly comparable with another.

03

Remove artifact before calculating

Eye blinks, eye movement, jaw and forehead muscle, movement, heartbeat, poor contact, and electrical noise can imitate brain activity. The report should state how much data was rejected and how much clean data remained.

04

Read the raw EEG first

A trained clinician reviews the ordinary waveform for epileptiform discharges, slowing, asymmetry, sleep features, and artifact before treating software maps or numbers as clinically meaningful.

05

Calculate the named measures

Software may calculate absolute and relative delta, theta, alpha, or beta power, peak alpha frequency, asymmetry, coherence, phase, ratios, or other features. Keep the software's exact frequency boundaries and method.

06

Compare with the exact database

A z-score commonly subtracts the database's age-dependent mean from the person's value and divides by the database standard deviation. Save the database version, sample, age range, sex handling, device, montage, reference, and correction for testing many values.

04

How do you read qEEG brain maps and z-scores?

Start with recording quality and the conventional EEG interpretation. Then read the state, montage, reference, clean duration, software, database, power or connectivity measure, z-score, and multiple-comparison method. Finish with symptoms, medicines, sleep, age, the clinician's conclusion, and the decision that changed.

The recording or comparison is not usable enough

The map may be unreliable if there is too little clean data, drowsiness, eye or muscle noise, movement, or poor electrode contact. A nonstandard electrode setup, unclear reference, missing raw review, or a poor database match can also cause problems. Ask whether a repeat would help.

No clear abnormality appears under the named method

The conventional EEG is unremarkable and no clear qEEG outlier appears under the named method

This recording didn't show a clear problem. Conditions or symptoms that did not appear during the test may still need other checks.

One or more values are statistical outliers

One or more power, asymmetry, coherence, phase, ratio, or frequency values fall outside the database expectation

This is a statistical difference from the named database. Read its z-score, location, band, direction, repeatability, artifact check, raw waveform, database match, and multiple-comparison method. It isn't a diagnosis, a damaged-brain percentage, or proof that the highlighted area caused symptoms.

The raw EEG or a repeatable feature needs clinical follow-up

The raw EEG shows a clinically important finding or a repeatable quantitative change supports a defined question

A trained clinician should name the conventional EEG finding, the quantitative feature, and what it changes. Epileptiform discharges, focal or generalized slowing, or another traditional EEG abnormality may lead to a conventional neurological workup. The color map shouldn't replace that explanation.

A z-score of 2 does not mean a 2 percent chance of disease

It means that one value is two standard deviations from one database model's fitted mean. The meaning changes with direction, channel, frequency band, state, age match, device, reference, artifact, and how many values were tested. It does not show damaged-brain percentage, disease probability, or treatment need.

See research details

Electrode count, reference database, z-score method, false-positive risk, FDA-cleared scope, developmental norms, and country access all affect interpretation. Research on ADHD, concussion, and brain fog does not remove those limits.

SourceA standard clinical EEG and a research qEEG montage are not the same number ContextACNS minimum technical guidance says a standard clinical EEG should use all 21 electrodes and placements recommended by the IFCN under the 10-20 system. The HarMNqEEG research project analyzed 19 scalp channels. Special clinical circumstances can justify a smaller montage, but it is not a complete standard clinical EEG.

Save the electrode count, names, montage, reference, sampling rate, filters, impedance method, recording states, and clean duration. Do not compare a reduced commercial map with a full clinical recording as though the methods were identical.

SourceElectrode contact matters before any map is calculated ContextACNS guidance says electrode impedance should ordinarily not be above 5 kOhm or below 100 Ohm. Modern amplifiers may sometimes record satisfactorily up to 10 kOhm, but balanced low impedances remain important. These are technical contact values, not brain-health scores.

Ask whether contact quality was acceptable and balanced throughout the recording. Do not interpret a focal hot spot before the technologist and reader have excluded poor contact and electrical noise.

SourceA z-score shows distance from one database, not the chance of disease ContextLi 2022 describes z as the person's value minus the age-dependent reference mean, divided by the age-dependent standard deviation. A z-score of 2 is two database standard deviations from the fitted mean. It is not a 2 percent risk, a diagnosis, or a treatment target.

Keep the sign, measure, electrode, frequency band, state, database, and uncertainty. Ask about the number of comparisons and the correction or validation method before focusing on the most colorful value.

SourceThe large multinational norm shows why method matching matters ContextHarMNqEEG began with 1,564 EEGs from 9 countries and 14 devices. One normative calculation removed 191 records during outlier and quality checks, leaving 1,373. It used at least 1 minute of artifact-free eyes-closed resting EEG and found that age, country, device, and study batch could affect measures.

Ask whether the person's age, device, reference, state, and processing match the database. A large sample does not erase batch effects or make a mismatched comparison accurate for that person.

SourceTesting many values can create apparently abnormal colors ContextA retired 1997 AAN and ACNS statement warned that, in normal controls, false abnormalities averaged about 5 percent of tests and could reach 15 to 20 percent in an individual when many statistical comparisons were made. The document is historical and retired, but the multiple-testing problem remains a basic statistical issue.

Do not use the old percentages as a modern device-accuracy estimate. Ask the current service how it controls false discovery, validates features, and confirms that a flagged region is not artifact or a database mismatch.

SourceThe FDA-cleared NEBA device is narrow, not generic brain-map approval ContextFDA De Novo K112711 describes a prescription aid for ages 6 through 17 after clinical evaluation. It uses at least 30 seconds of still recording from one Cz electrode, with Fz ground, linked-ear reference, EOG, and different theta-to-beta thresholds for ages 6.00 to 11.99 and 12.00 to 17.99.

Check the device name, product code, age, intended use, electrodes, and exact claim. Do not transfer one device's clearance to a 19-channel or 21-electrode map, an adult, or a diagnosis and treatment claim it was not cleared to make.

SourceCurrent ADHD guidance blocks stand-alone theta-to-beta diagnosis ContextThe AAN advisory on EEG theta-to-beta ratio, reaffirmed October 18, 2025, says clinicians should tell families it should not replace a standard clinical evaluation or confirm ADHD outside research. In the reviewed evidence, false-positive ADHD diagnosis occurred in nearly 6 to 15 percent.

Use history, onset, impairment in more than one setting, observer information, development, sleep, learning, mood, substances, medicines, and other causes. A ratio or color map cannot supply those facts.

SourceThe 2021 concussion guideline says qEEG remains investigational ContextTenney 2021 found class III evidence and concluded that available evidence does not support qEEG for diagnosing mild traumatic brain injury, distinguishing it from other diagnoses, or detecting it in the presence of central-nervous-system medicine.

Do not use an abnormal or expected map to confirm or exclude concussion. Keep the injury history, examination, danger signs, symptom course, function, medicines, sleep, vestibular findings, and conventional imaging or testing decisions separate.

SourceBrain-fog-related conditions do not share one established pattern ContextSilva-Passadouro 2024 found 2,510 records and included 17 studies of fibromyalgia, ME/CFS, or Long COVID. Fibromyalgia studies tended toward less low-frequency delta, theta, and alpha power with more beta; ME/CFS patterns differed; the few Long COVID studies were mixed.

Treat these as group-level research findings. They do not create a Long COVID, ME/CFS, fibromyalgia, or brain-fog diagnostic signature and do not tell an individual which treatment to buy.

SourceA 2025 youth model performed well inside its research design ContextDong 2025 built a multi-site normative framework from 1,212 young people and reported intraclass correlation above 0.9 for its research model. The paper also addressed site and montage batch effects. This is methods progress, not validation of every commercial map or diagnosis.

For a child or teenager, ask whether the task, montage, processing, age model, developmental stage, and clinical use match the evidence. Do not replace pediatric assessment with a research reliability number.

SourceThe 2026 pediatric review found large method differences ContextGray 2026 screened more than 650 papers and extracted methods from 56 pediatric resting-EEG studies. Acquisition, preprocessing, frequency-band definitions, and handling of multiple comparisons varied substantially, and the variation was not explained by age, publication year, or paradigm.

Require the report to name the child's exact method and age comparison. A pediatric-looking color map cannot be interpreted without developmental and technical context.

SourceA strong 2026 depression estimate still does not create a routine diagnostic test ContextJing 2026 pooled 18 studies with 58 results and reported sensitivity 0.939, specificity 0.898, and a diagnostic odds ratio of 137.493. The authors also identified case-control selection bias, small samples, and unexplained clinical and statistical heterogeneity.

Keep both the numbers and the limits. This supports further diagnostic research, not using one qEEG report to diagnose depression, choose an antidepressant, or explain brain fog.

SourceCountry access and coverage are not the same as clinical validity ContextIn the United States, check the exact FDA device and insurer policy. UK EEG is usually a specialist service, while commercial qEEG availability varies. Australia's MBS item 11005 for prolonged 3-to-24-hour EEG explicitly excludes quantitative topographic mapping using neurometrics or similar devices.

Ask for the service name, clinician credentials, device, billing code, prior authorization, total cost, and what part is clinical EEG versus optional quantitative mapping. Insurance coverage doesn't prove a diagnostic claim, and lack of coverage doesn't by itself settle the science.

05

Make the appointment useful without trying to change the trace

Practising the task cannot make the assessment more trustworthy. You can still improve the information available at the appointment and reduce how much the symptoms disrupt daily life while assessment continues.

Keep the recording conditions ordinary and written down

Follow the clinic's plan for sleep, meals, caffeine, nicotine, and medicines. Write the actual hours slept, last caffeine, food, illness, pain, migraine, and medicine times instead of trying to create a better or worse trace.

Clean, dry hair on arrival

Wash and dry your hair. Skip gel, wax, oil, and spray. Ask about braids, extensions, wigs, scalp sensitivity, skin problems, or a religious head covering before the visit. Clean contact can reduce delay and recording noise.

Bring two daily-life examples with the qEEG report

Write what happened, how long it lasted, what you were doing, what became impossible, and how long recovery took. Include sleep, upright posture, exertion, meals, migraine, sensory load, medicine timing, and menstrual or postpartum timing when relevant.

Ask for the raw EEG and the method page

Request the clinician's ordinary EEG interpretation plus the quantitative report, electrode montage, state, artifact-free duration, software, database, z-score method, and limits. Ask which single clinical decision changed because of the result.

Reduce the cost of the symptom now

Use written steps, reminders, fewer simultaneous demands, planned breaks, a quieter work area, and help with driving or childcare when episodes make those tasks unsafe. These practical supports don't claim that a brainwave is now normal.

Check the treatment claim before paying

If a clinic suggests neurofeedback, supplements, medicine changes, or repeat maps, ask why. Request the controlled evidence, total cost, expected benefit, possible harm, other choices, stopping rule, and an independent review.

Where preparation and self-support stop

Do not stop prescribed medicine, lose sleep on purpose, trigger symptoms, or buy treatment to chase a color or z-score. Get direct medical care for a first or repeat seizure, sudden weakness or numbness, or new trouble speaking or seeing. A severe sudden headache, marked confusion, or reduced consciousness also needs direct care.

06

What should you keep with a qEEG report?

Keep these together

  • Referral question, symptom or event under review, and decision the test was meant to change
  • Clinical EEG, qEEG analysis, commercial brain map, or named cleared device
  • Clinic, qualified reader, device, amplifier, software, database, and version
  • Date, start time, eyes-open or eyes-closed state, task, drowsiness, sleep, and total recording duration
  • Montage, every electrode used, reference, sampling rate, filters, contact or impedance notes, and activation procedures
  • Artifact methods, rejected segments, remaining artifact-free minutes, movement, eye, muscle, heart, and electrical-noise notes
  • Conventional raw EEG interpretation, including epileptiform activity, focal or generalized slowing, asymmetry, or an unremarkable study
  • Absolute and relative power, exact delta, theta, alpha, and beta boundaries, and peak alpha frequency
  • Asymmetry, coherence, phase, ratios, connectivity, or other derived measures actually reported
  • Every highlighted value, z-score, sign, channel, band, state, unit, database comparison, and multiple-comparison method
  • Age and sex model, country and device match, normative sample, exclusions, and report limits
  • Sleep, food, caffeine, nicotine, alcohol, cannabis, illness, fever, pain, migraine, anxiety, and medicine timing
  • Pregnancy or postpartum timing, hearing, vision, scalp, movement, language, sensory, and access context
  • Clinician interpretation, competing explanations, conventional follow-up, treatment recommendation, cost, expected benefit, risks, and stopping rule
  • For a repeat, the reason, interval, same or changed method, clinical event, treatment, symptoms, function, and decision it will test

Question for the visit

“Which part came from the raw clinical EEG, which part came from the quantitative comparison, how well did the method and database fit me, what could be artifact or normal variation, and what clinical decision changes because of this report?”
07

Sources for Quantitative EEG (qEEG) Brain Mapping

01
www.acns.org
02
American Clinical Neurophysiology Society: minimum technical EEG requirements

Standard 21-electrode clinical recording, 10-20 placement, impedance, montage, and technical quality

03
Retired 1997 AAN and ACNS qEEG statement

Historical raw-EEG-first rule, trained interpretation, multiple-testing warning, and limited established adjunct uses

04
American Academy of Neurology ADHD EEG advisory, reaffirmed 2025

Theta-to-beta ratio diagnostic limits, false positives, and standard clinical-assessment requirement

05
US FDA De Novo K112711: NEBA system

Narrow prescription-device use, ages 6 to 17, one-channel method, duration, thresholds, and limitations

06
NHS: electroencephalogram

Current preparation, clean hair, ordinary food and drink, basic procedure, result review, and sleep-deprived safety

07
Healthdirect Australia: EEG, reviewed 2025

Non-invasive recording, usual duration, hair preparation, medicine review, and sleep-deprived safety

08
Australian Medicare Benefits Schedule item 11005

Prolonged EEG item and explicit exclusion of quantitative topographic neurometric mapping

09
Nuwer, 1997, PMID 9222209, retired statement

Historical qEEG evidence classification, raw EEG, trained interpretation, and false-abnormality warning

10
Gloss et al., 2016, PMID 27760867

AAN ADHD theta-to-beta ratio advisory, false-positive range, and diagnostic boundary

11
Tenney et al., 2021, PMID 34038930

ACNS guideline against qEEG diagnosis of mild traumatic brain injury

12
Li et al., 2022, PMID 35398285

HarMNqEEG sample, countries, devices, channels, clean duration, z-score method, age, and batch effects

13
Silva-Passadouro et al., 2024, PMID 38772083

Seventeen-study fibromyalgia, ME/CFS, and Long COVID systematic review and mixed-results limit

14
Dong et al., 2025, PMID 40946930

1,212-person youth normative framework, multi-site and montage batch effects, and research reliability

15
Gray et al., 2026, PMID 42176402

Pediatric resting-EEG methods review, screened and included counts, and method variability

16
Jing et al., 2026, PMID 42334653

Depression diagnostic-accuracy meta-analysis, pooled estimates, bias, small samples, and heterogeneity

See each claim's sources

limitation

The 2021 ACNS guideline does not support qEEG for diagnosing mild traumatic brain injury or distinguishing it from other diagnoses and calls this use investigational.

indication

FDA De Novo K112711 covers a prescription one-channel Cz theta-to-beta ratio aid for ages 6 through 17 after a clinical evaluation, not generic diagnostic brain mapping.

context

HarMNqEEG began with 1,564 EEGs from 9 countries and 14 devices and used 19 channels and at least 1 minute of artifact-free eyes-closed rest in its harmonized norm work.

interpretation

A qEEG z-score expresses distance from the named age-dependent database mean in database standard-deviation units, not disease probability.

limitation

A 2024 systematic review included 17 fibromyalgia, ME/CFS, or Long COVID qEEG studies and found differing or mixed group patterns rather than one clinical brain-fog signature.

context

A 2025 normative framework used 1,212 young people and reported research-model reliability above 0.9 while addressing site and montage batch effects.

limitation

A 2026 pediatric review screened more than 650 papers, extracted 56, and found substantial variation in acquisition, preprocessing, frequency bands, and multiple-comparison handling.

interpretation

A 2026 depression meta-analysis pooled 18 studies and 58 results but retained case-control bias, small samples, and unexplained heterogeneity despite high pooled accuracy estimates.