eduKate Learning Manual
Science | Veterinary World
Describe the Event → Localise the Brain Question → Record Cortical Electrical Activity → Separate Ictal From Interictal Evidence → Check Artefact and Protocol → Integrate With Clinical Neurology → Reassess Over Time

Veterinary EEG

Why a Normal Interictal EEG Does Not Rule Out Epilepsy

Wait, What? A Brain Can Be Epileptic Even When the EEG Looks Normal Between Seizures

Electroencephalography can feel like the perfect epilepsy test. If seizures are electrical storms in the brain, surely an electrical recording should either show the storm or prove that no storm exists.

That is not how intermittent biology behaves.

EEG records electrical activity only during the time and from the regions sampled. An animal with epilepsy may have abnormal discharges during a seizure but little or no detectable epileptiform activity during a short interictal recording. Deep or focal activity may also be poorly represented at the scalp. Sedation, anaesthesia, electrode placement, skull anatomy, muscle artefact and recording duration can all change what becomes visible.

normal interictal EEG ≠ epilepsy excluded.

The Scientific Job

This page owns one Veterinary World job:

How should veterinarians interpret canine and feline EEG recordings as time-limited samples of cortical electrical activity, distinguishing ictal, interictal and artefactual patterns without mistaking a normal recording for proof that epilepsy is absent?

Veterinary Neurological Localisation retains the broader anatomical localisation of brain, spinal cord, nerve and muscle disease. Veterinary Syncope vs Seizure retains event differentiation. This page owns the narrower job of recording and interpreting cerebral electrical activity.

Quick Answer

EEG is the patient-side test that directly records cortical electrical activity and can provide objective evidence that an event is epileptic. It is particularly powerful when an event is recorded during the EEG. Interictal EEG is less sensitive: epileptiform discharges can be intermittent, focal or absent during a short sample. A normal interictal study therefore reduces little unless the protocol, duration and biological question made the abnormality reasonably likely to appear.

A 2023 survey of veterinary neurologists found that EEG remains underused in dogs despite being the direct test of seizure-related electrical activity. A 2024 review highlighted continuing variation in canine electrode arrays and the need for better standardisation, while a 2025 systematic review of non-invasive canine EEG showed growing use but substantial methodological heterogeneity.

Explore 2023 Survey — EEG Usage and Techniques in Dogs →

Explore 2024 Review — EEG Electrode Arrays in Dogs →

Explore 2025 Systematic Review — Non-Invasive Canine EEG →

Primary Entry — EEG Is a Time Sample of a Dynamic Brain

The cortex contains enormous populations of neurons whose postsynaptic electrical activity can sum into voltage differences measurable at the scalp. EEG electrodes do not read individual neurons. They sample the combined electrical fields generated by many synchronised cortical cells.

That gives EEG extraordinary temporal resolution. Millisecond-scale changes are visible. But it also creates limitations: the signal is strongest for superficial, synchronised cortical activity and weaker for deep, small or poorly oriented generators.

Part 1 — Ictal and Interictal Are Different Biological States

An ictal EEG records the brain during an actual seizure. Rhythmic or evolving electrical patterns can provide direct evidence of epileptic activity.

An interictal EEG records between seizures. It may show spikes, sharp waves or other epileptiform discharges that support epilepsy, but it may also look ordinary because the abnormal network is not discharging visibly during that sample.

epilepsy is a disorder of recurrence; one interictal sample may miss a recurrent process.

Part 2 — A Short Recording Can Miss a Rare Electrical Event

Suppose interictal epileptiform discharges occur only a few times each hour. A brief recording may contain none. The EEG can be technically excellent and still miss the event because the sampling window was too short.

Older canine work under propofol restraint found epileptiform discharges in only a minority of dogs with epilepsy during interictal recordings. The key lesson is not that EEG is useless. It is that sensitivity depends strongly on biological timing and protocol.

Explore Canine Interictal EEG Under Propofol Restraint →

Part 3 — Event Capture Is More Powerful Than Event Memory

Owners may describe trembling, collapse, staring, paddling or unresponsiveness. Those observations matter, but several non-epileptic events can imitate seizure behaviour.

If the suspected episode occurs while EEG is recording and the trace shows an evolving epileptic pattern at the same time, the evidence becomes much stronger. If the episode occurs with no corresponding cerebral electrical change, alternative diagnoses become more important, although technical limitations must still be considered.

Part 4 — Scalp EEG Does Not See Every Brain Region Equally

Electrical fields weaken and spread as they travel through cerebrospinal fluid, skull, muscle and skin. Deep temporal or subcortical seizure activity can therefore be harder to detect from scalp electrodes than widespread superficial cortical activity.

This means a normal scalp EEG cannot be promoted into the statement “the entire brain was electrically normal”. It means no diagnostically convincing abnormality was recorded through that montage under those conditions.

Part 5 — Electrode Placement Changes What the Recording Can See

Human EEG uses highly standardised electrode systems. Canine EEG still uses multiple arrays and placement approaches across centres. The 2024 review of canine electrode arrays highlighted this lack of complete standardisation and the importance of validating how well different montages localise cortical activity.

A sparse montage can miss or blur focal activity that a denser, well-positioned array might reveal.

Secondary Deepening — Artefact Is Often Bigger Than the Brain Signal

EEG signals are measured in microvolts. Jaw movement, eye movement, heart activity, respiration, electrode movement and environmental electrical noise can all produce signals larger than the cortical activity being studied.

Muscle artefact is particularly important in awake dogs because scalp and temporalis muscles can generate fast electrical activity that resembles or obscures cerebral rhythms. The interpreter must therefore ask whether the waveform follows brain physiology or the animal’s movement.

Part 6 — Sedation Solves One Problem and Creates Another

Keeping an animal still improves electrode contact and reduces movement artefact. Sedation or anaesthesia can help achieve that. Yet those drugs also change brain electrical activity.

The result is a trade-off: cleaner recording versus altered physiology. Interpretation is strongest when the drug, dose context and expected EEG effects are documented rather than hidden.

Part 7 — Spikes Are Evidence, Not a Complete Diagnosis

Spikes and sharp waves can support increased cortical excitability, especially when morphology, field distribution and clinical context fit epilepsy. But isolated sharp transients can also arise from artefact, normal variants or non-epileptic brain disease.

The 2023 visual review of periodic discharges in veterinary EEG demonstrates how some abnormal patterns can sit near the boundary between ictal, interictal and encephalopathic activity. Pattern recognition therefore requires context rather than one dramatic waveform.

Explore Review — Periodic Discharges in Veterinary EEG →

Part 8 — Background Activity Matters Too

EEG interpretation is not only a search for spikes. Background rhythm, symmetry, reactivity, slowing and organisation can provide evidence about diffuse or focal cerebral dysfunction.

Generalised slowing may support encephalopathy or sedative effect rather than epilepsy specifically. Focal slowing can suggest local cerebral dysfunction but does not identify the exact lesion.

JC Deepening — EEG Is a Sampling-Theory Problem

There are two dimensions of sampling:

• time: did the abnormal event occur during the recording?
• space: did the electrode array sample the cortical region strongly enough to detect it?

A normal EEG therefore has different meaning depending on recording duration, montage, disease frequency and lesion depth.

negative evidence becomes stronger only when the test had a fair opportunity to observe the abnormality.

Part 9 — Continuous EEG Changes the Question

Longer recordings increase the probability of capturing sporadic abnormalities and can reveal non-convulsive seizures or seizure clusters that brief examinations miss.

Research into wireless and long-duration canine EEG aims to record dogs in more natural environments, where spontaneous seizures can be captured without relying entirely on owner diaries. This could substantially improve event classification and seizure counting.

Explore Continuous Wireless EEG in Behaving Dogs →

Part 10 — EEG and MRI Answer Different Questions

MRI shows structure. EEG shows electrical function across time.

A dog can have normal MRI with idiopathic epilepsy and abnormal EEG. Another can have a structural brain lesion on MRI but no epileptiform discharge during a short EEG. The tests are complementary because anatomy and electrical activity are different dimensions.

Part 11 — A Normal Interictal EEG Can Still Be Useful

Normal does not mean worthless. A technically strong normal recording can reduce concern for some persistent electrical abnormalities, provide a baseline for later comparison and reveal that a recorded event lacked an ictal correlate.

The error is not in calling the recording normal. The error is turning “normal during this sample” into “epilepsy impossible”.

Part 12 — Standardisation Is the Next Major Veterinary EEG Problem

The recent literature repeatedly identifies variability in electrode arrays, preparation, sedation, recording duration and interpretation. Without harmonisation, two laboratories can record the same biological phenomenon differently.

World-class veterinary EEG therefore needs not only better hardware, but common protocols and transparent metadata about how each trace was produced.

How Do We Know?

Veterinary evidence includes clinical EEG surveys, technique reviews, interictal epilepsy studies, visual analyses of pathological waveforms and newer systematic reviews of non-invasive canine EEG. Together they support EEG as the direct electrical test for epileptic activity while also showing major limitations in sensitivity, standardisation and event capture when recordings are brief or interictal.

Observation vs Inference

• Observation: a dog has a normal 20-minute interictal EEG.
• Inference: no convincing epileptiform abnormality was recorded during that interval; epilepsy remains possible.
• Observation: a stereotyped event occurs during EEG with an evolving rhythmic cortical discharge.
• Inference: the event is strongly supported as epileptic.
• Observation: sharp transients appear only with jaw movement.
• Inference: muscle artefact becomes more likely than cortical epileptiform activity.
• Observation: background slowing is diffuse without epileptiform discharge.
• Inference: diffuse cerebral dysfunction, medication effect or metabolic encephalopathy may be more relevant than epilepsy alone.

Evidence Boundaries

• normal interictal EEG ≠ epilepsy excluded.
• one spike ≠ seizure disorder proven.
• scalp EEG ≠ complete access to every brain region.
• sedated EEG ≠ untreated waking physiology.
• background slowing ≠ one specific diagnosis.
• absence during short recording ≠ absence across the day.
• EEG abnormality ≠ structural lesion identified.
• electrical diagnosis ≠ treatment instruction.

Common Misconceptions

Misconception	Better model
Normal EEG means no epilepsy.	Interictal epileptiform activity can be intermittent and may be missed.
Any spike is epilepsy.	Artefact and other cerebral abnormalities can mimic or create sharp transients.
MRI replaces EEG.	MRI measures structure; EEG measures electrical function over time.
A brief EEG samples the whole day.	It is a narrow time window whose sensitivity depends on event frequency.

Unfamiliar Transfer

Dog A has recurrent stereotyped seizures but a normal 30-minute EEG. Dog B has a strange collapse captured during EEG without a cerebral electrical correlate. Dog C has frequent sharp waves but also severe jaw-muscle artefact. Dog D has diffuse slowing and metabolic encephalopathy.

A strong learner does not reduce all four to “EEG positive or negative”. The learner asks what was sampled, whether the event was captured, whether the waveform is physiological or artefactual, and which layer of the neurological problem remains unresolved.

Checkpoint Questions

1. What does EEG physically measure?
2. Why can an animal with epilepsy have a normal interictal EEG?
3. What is the difference between ictal and interictal recording?
4. Why does recording duration matter?
5. How can electrode placement affect sensitivity?
6. Why is muscle artefact difficult in dogs?
7. How can sedation alter interpretation?
8. Why can background slowing matter even without spikes?
9. Why are EEG and MRI complementary?
10. What makes a negative EEG more convincing?

Answer key

1. Summed electrical activity generated by populations of cortical neurons.
2. Epileptiform discharges can be intermittent, focal, deep or absent during the sampled window.
3. Ictal is during a seizure; interictal is between seizures.
4. Longer recordings increase the chance of capturing sporadic abnormalities.
5. Different montages sample cortical regions with different spatial sensitivity.
6. Muscle signals can be much larger than cortical microvolt activity.
7. Drugs reduce movement but also change cerebral electrical patterns.
8. It can support diffuse or focal cerebral dysfunction even when epilepsy is not directly demonstrated.
9. One measures electrical function; the other measures anatomy.
10. When the protocol, duration and montage made the suspected abnormality reasonably likely to be detected.

Edge Science — Can Continuous Home EEG Turn Seizure Diaries Into Measured Events?

Wireless, subscalp and longer-duration canine EEG systems are moving the field toward recordings in awake animals living ordinary lives. That could reveal seizures that owners miss and distinguish behavioural events from electrical seizures without relying entirely on memory.

The challenge is data volume and false alarms. Automated seizure detection must handle movement, scratching, sleep transitions and breed-specific skull geometry. A trustworthy system should preserve raw waveform segments, confidence and recording quality rather than output a hidden seizure count.

Veterinary World Direction Graph

Veterinary EEG → paroxysmal event → neurological localisation → electrode montage → recording state → ictal/interictal/background pattern → artefact check → MRI/metabolic context → event correlation → serial/long-term monitoring.

Research Sources and Further Reading

• Survey of Electroencephalography Usage and Techniques for Dogs
• 2024 Review — Electroencephalography Electrode Arrays in Dogs
• 2025 Systematic Review — Non-Invasive Canine EEG
• Periodic Discharges in Veterinary EEG — Visual Review
• Continuous Wireless EEG in Unsedated Behaving Dogs

Educational boundary: Recurrent seizures, prolonged seizures, cluster seizures or sudden neurological deterioration require veterinary assessment. This manual explains EEG evidence only and does not provide anti-seizure drug selection, dosing or emergency treatment instructions.

Teaching Guide for Parents, Tutors and Teachers

For the people who teach because somebody depends on them.

Ask: “If lightning strikes one night a month, does watching the sky for ten quiet minutes prove there is never lightning?” The observation can be completely correct and still be too short to answer the larger question.

define the event → sample the brain → ask whether the event occurred during the sample → check artefact → integrate anatomy and physiology → preserve uncertainty after a negative recording.

The mastery target is a learner who understands that a negative measurement is only as powerful as the opportunity the test had to witness the thing being sought.