Evoked Potentials
Evoked Potentials
What Do You Need to Know?
- General principles — signal averaging extracts time-locked CNS responses from background noise; latency = myelin integrity, amplitude = axonal integrity
- VEP — P100 waveform (striate cortex); prolonged P100 latency = optic neuritis/MS; most sensitive EP for MS
- BAEP — 5 waves (CN VIII to inferior colliculus); Wave V most robust; I-III interval = acoustic neuroma; absent waves beyond I = brain death
- SSEP — dorsal column pathway; N20 (upper) and P37 (lower) = cortical responses; bilateral absent N20 post-cardiac arrest = poor prognosis (most reliable prognostic tool)
- MEP — transcranial magnetic stimulation of motor cortex; CMCT = cortical latency minus spinal latency; corticospinal tract assessment
- Intraoperative monitoring — SSEP for posterior columns, MEP for anterior cord, BAEP for CN VIII; alarm = 50% amplitude drop or 10% latency increase
- EPs in MS — VEP most sensitive (85–90%), SSEP (50–70%), BAEP (50–60%); detect subclinical demyelinating lesions
Overview
General Principles
- Evoked potentials (EPs) — measure CNS conduction along specific sensory or motor pathways in response to a defined stimulus
- Signal averaging — hundreds to thousands of repetitions are averaged to extract the time-locked response from random background EEG noise
- Latency — reflects myelin integrity (prolonged in demyelination)
- Amplitude — reflects axonal integrity and number of functioning fibers (reduced in axonal loss)
- Primary clinical uses — detect subclinical lesions (especially in MS), intraoperative monitoring, prognostication
- EPs test the entire pathway from stimulus to cortex — can localize lesion to peripheral, brainstem, or cortical segments
Key Terminology
- Absolute latency — time from stimulus to a specific peak
- Interpeak latency (IPL) — time between two peaks; localizes the segment of pathway involved
- Central conduction time (CCT) — transit time through the CNS (excludes peripheral segment)
- Waveforms named by polarity and latency: N = negative, P = positive, number = approximate latency in ms (e.g., N20 = negative peak at ~20 ms)
Board Pearl
Latency = myelin; amplitude = axons. Prolonged latency with preserved amplitude suggests demyelination. Reduced amplitude with normal latency suggests axonal loss or conduction block. This principle applies to all EP modalities.
Visual Evoked Potentials (VEP)
Technique and Waveforms
- Stimulus — pattern-reversal checkerboard (alternating black/white squares on a screen); each eye tested separately
- Recording — active electrode at Oz (occipital midline), referenced to Fz (midfrontal)
- Key waveform: P100 — positive peak at ~100 ms; generated by striate cortex (V1)
- Full-field stimulation — entire visual field of one eye; detects pre-chiasmal lesions (optic nerve)
- Half-field stimulation — one hemifield at a time; localizes post-chiasmal lesions (optic tract, radiation)
Abnormalities
| Abnormality | Mechanism | Clinical Significance |
|---|---|---|
| Prolonged P100 latency | Demyelination of optic nerve | Optic neuritis, MS (most common cause) |
| Reduced P100 amplitude | Axonal loss in optic nerve | Compressive optic neuropathy, severe optic neuritis, ischemic optic neuropathy |
| Absent P100 | Complete conduction failure | Severe optic nerve damage, technical issue (check visual acuity) |
| Asymmetric half-field responses | Post-chiasmal lesion | Optic tract or radiation lesion; homonymous pattern |
Board Pearl
VEP is the most sensitive EP for detecting MS. P100 latency remains prolonged even after clinical recovery from optic neuritis — it serves as a permanent "fingerprint" of prior demyelination. A prolonged P100 in a clinically unaffected eye supports dissemination in space.
Brainstem Auditory Evoked Potentials (BAEP)
Technique
- Stimulus — monaural clicks delivered via earphones (alternating polarity); contralateral ear receives white noise masking
- Recording — vertex (Cz) referenced to ipsilateral ear (A1 or A2)
- 5 waveforms generated within 10 ms of stimulus
BAEP Waves — Generators
| Wave | Approximate Latency | Generator | Anatomic Level |
|---|---|---|---|
| I | ~1.5 ms | Distal CN VIII (cochlear nerve) | Inner ear / distal nerve |
| II | ~2.5 ms | Proximal CN VIII / cochlear nucleus | Pontomedullary junction |
| III | ~3.5 ms | Superior olivary complex | Lower pons |
| IV | ~4.5 ms | Lateral lemniscus | Upper pons |
| V | ~5.5 ms | Inferior colliculus | Midbrain |
Memory Aid: BAEP Wave Generators
- "E-COSLIn" — Eighth nerve (I), Cochlear nucleus (II), Olivary complex (III), Lateral lemniscus (IV), Inferior colliculus (V)
- Wave V is the most robust and last wave to disappear — it is the most clinically important waveform
Interpeak Latencies
| Interpeak Interval | Segment Tested | Abnormal In |
|---|---|---|
| I–III | CN VIII to lower pons | Acoustic neuroma (vestibular schwannoma), CPA tumors |
| III–V | Lower pons to midbrain | MS (brainstem demyelination), brainstem stroke |
| I–V | Entire central auditory pathway | Any brainstem lesion; prolonged in MS |
Clinical Applications
- Acoustic neuroma — prolonged I–III IPL or absent waves beyond Wave I
- MS — prolonged III–V or I–V IPL (central demyelination)
- Intraoperative monitoring — posterior fossa surgery, CN VIII preservation
- Brain death — absent all waves, or only Wave I present (peripheral response without brainstem conduction)
- Infant hearing assessment — objective hearing threshold estimation in neonates (no patient cooperation needed)
Clinical Pearl
In acoustic neuroma screening, BAEP has been largely replaced by MRI with gadolinium. However, BAEP remains valuable intraoperatively to monitor CN VIII function during posterior fossa and cerebellopontine angle surgery.
Somatosensory Evoked Potentials (SSEP)
Technique
- Stimulus — electrical stimulation of peripheral nerve (square-wave pulse at motor threshold)
- Upper extremity — median nerve at wrist (most common)
- Lower extremity — posterior tibial nerve at ankle (most common)
- Pathway tested — peripheral nerve → dorsal columns → medial lemniscus → thalamus (VPL) → primary sensory cortex
Upper Extremity SSEP Waveforms (Median Nerve)
| Waveform | Approximate Latency | Generator | Recording Site |
|---|---|---|---|
| N9 | ~9 ms | Brachial plexus (Erb's point) | Erb's point |
| N13 | ~13 ms | Cervical cord / dorsal horn (cervical spine) | Posterior neck (C5 spinous process) |
| P14 | ~14 ms | Medial lemniscus / cervicomedullary junction | Scalp (far-field) |
| N20 | ~20 ms | Primary somatosensory cortex (contralateral) | Scalp (C3' or C4') |
Lower Extremity SSEP Waveforms (Posterior Tibial Nerve)
| Waveform | Approximate Latency | Generator | Recording Site |
|---|---|---|---|
| LP (lumbar potential) | ~20 ms | Cauda equina / lumbar cord | Lumbar spine |
| N34 | ~34 ms | Brainstem (medial lemniscus) | Scalp (far-field) |
| P37 | ~37 ms | Primary somatosensory cortex | Scalp (Cz') |
Clinical Applications
- MS — central conduction delay (prolonged N13–N20 or LP–P37 interpeak latency); subclinical spinal cord/brainstem lesions
- Intraoperative spinal cord monitoring — monitors posterior column (dorsal column) function during scoliosis correction, spinal tumor surgery
- Post-cardiac arrest prognostication — bilateral absent cortical N20 at 24–72 hours = poor neurologic prognosis
- Myelopathy evaluation — detect subclinical dorsal column dysfunction in cervical spondylotic myelopathy, B12 deficiency
Board Pearl
Bilateral absent N20 on SSEP at 24–72 hours post-cardiac arrest is the most reliable prognostic indicator of poor neurologic outcome (false positive rate <1%). It is the single most specific test for predicting failure to regain consciousness after anoxic brain injury. This is a favorite board question.
Motor Evoked Potentials (MEP)
Technique
- Stimulus — transcranial magnetic stimulation (TMS) or transcranial electrical stimulation (TES) over motor cortex
- Recording — compound muscle action potential (CMAP) from target muscle (e.g., hand intrinsics, tibialis anterior)
- Pathway tested — motor cortex → corticospinal tract → anterior horn cell → peripheral nerve → muscle
Central Motor Conduction Time (CMCT)
- CMCT = cortical MEP latency − spinal (root) MEP latency
- Isolates the central segment (cortex to anterior horn cell)
- Normal CMCT — ~6–8 ms (upper extremity), ~12–16 ms (lower extremity)
- Prolonged CMCT → corticospinal tract demyelination (MS, hereditary spastic paraplegia) or degeneration (ALS)
Clinical Applications
| Indication | Finding | Significance |
|---|---|---|
| MS | Prolonged CMCT | Corticospinal tract demyelination |
| ALS | Prolonged CMCT, reduced amplitude, absent responses | Upper motor neuron involvement (supports diagnosis) |
| Intraoperative monitoring | Amplitude/latency changes during surgery | Anterior cord / corticospinal tract ischemia detection |
| Myelopathy | Prolonged CMCT | Corticospinal tract compression |
Clinical Pearl
Intraoperative monitoring combines SSEP (posterior columns) with MEP (corticospinal tract) to cover both the anterior and posterior spinal cord. Isolated anterior cord ischemia (e.g., anterior spinal artery syndrome) may be missed by SSEP alone — MEP provides critical complementary monitoring.
EP Comparison Table
Master Comparison of All EP Modalities
| Feature | VEP | BAEP | SSEP | MEP |
|---|---|---|---|---|
| Stimulus | Pattern-reversal checkerboard | Monaural clicks | Electrical (median/tibial nerve) | TMS or TES over motor cortex |
| Pathway tested | Optic nerve → visual cortex | CN VIII → brainstem auditory pathway | Peripheral nerve → dorsal columns → sensory cortex | Motor cortex → corticospinal tract → muscle |
| Key waveform | P100 | Wave V (most robust) | N20 (upper) / P37 (lower) | CMAP from target muscle |
| Generator | Striate cortex (V1) | Inferior colliculus | Primary somatosensory cortex | Target muscle (via corticospinal tract) |
| Most common abnormality | Prolonged P100 latency | Prolonged I–III or III–V IPL | Prolonged central conduction time | Prolonged CMCT |
| Top clinical indication | Optic neuritis / MS detection | Acoustic neuroma, brainstem lesion, brain death | Spinal cord monitoring, post-arrest prognostication | Corticospinal tract assessment, intraoperative monitoring |
| Sensitivity for MS | 85–90% (highest) | 50–60% | 50–70% | Variable (50–70%) |
EPs in Multiple Sclerosis
Sensitivity by Modality
| EP Modality | Sensitivity in MS | What It Detects |
|---|---|---|
| VEP | 85–90% (with history of optic neuritis) | Optic nerve demyelination (even subclinical) |
| SSEP | 50–70% | Spinal cord / brainstem dorsal column demyelination |
| BAEP | 50–60% | Brainstem auditory pathway demyelination |
| MEP | 50–70% | Corticospinal tract demyelination |
Role in MS Diagnosis
- EPs detect subclinical lesions — objective evidence of demyelination in a pathway without clinical symptoms
- VEP showing prolonged P100 in a clinically unaffected eye supports dissemination in space
- In the 2017 McDonald criteria, VEP can be used as supportive evidence for optic nerve involvement
- Multimodal EPs (VEP + SSEP + BAEP) increase overall sensitivity for detecting MS lesions
- EPs are complementary to MRI — they detect functional demyelination that may not be visible on imaging
Board Pearl
VEP abnormality persists indefinitely after optic neuritis — even when visual acuity returns to normal, the P100 latency remains prolonged. This makes VEP valuable for proving prior optic neuritis in a patient who now has a normal exam. VEP is the most sensitive single EP modality for MS detection.
Intraoperative Monitoring
EP Modalities for Intraoperative Use
| Modality | Pathway Monitored | Surgery Type | Limitations |
|---|---|---|---|
| SSEP | Dorsal columns (posterior cord) | Scoliosis correction, spinal tumor, aortic surgery | Does NOT monitor anterior cord (motor tracts) |
| MEP | Corticospinal tract (anterior cord) | Spinal surgery, intracranial tumor near motor cortex | Sensitive to anesthetic agents (especially inhalational) |
| BAEP | CN VIII / brainstem auditory pathway | Posterior fossa surgery, CPA tumor resection | Only monitors auditory pathway |
| SSEP + MEP combined | Posterior + anterior spinal cord | Any spinal surgery (best practice) | Most comprehensive coverage |
Alarm Criteria
- Amplitude decrease ≥50% from baseline → significant (alert surgeon)
- Latency increase ≥10% from baseline → significant (alert surgeon)
- Complete loss of waveform → critical alert (immediate surgical response)
Alert Protocol
- Verify technical factors (electrode displacement, anesthetic changes, blood pressure, temperature)
- Notify surgeon immediately
- Consider reversible causes: hypotension, hypothermia, anesthetic depth change
- If changes persist after technical check → surgical cause likely → consider reversing recent surgical maneuver
Board Pearl
SSEP alone can miss anterior cord ischemia. The classic example is anterior spinal artery syndrome during aortic surgery — SSEPs may remain normal (dorsal columns spared) while the patient wakes with paraplegia (corticospinal tract infarcted). Adding MEP monitoring detects anterior cord compromise. Combined SSEP + MEP is the current standard of care.
Prognostication
Post-Cardiac Arrest
- Bilateral absent cortical N20 on SSEP at 24–72 hours → poor neurologic prognosis (persistent vegetative state or death)
- False positive rate <1% — this is the most specific and reliable prognostic tool after cardiac arrest
- SSEP is resistant to sedation and hypothermia, making it more reliable than clinical exam or EEG in the ICU setting
- Preserved N20 does NOT guarantee good outcome (low positive predictive value for recovery)
- Should be interpreted in conjunction with other prognostic tools (clinical exam, EEG, NSE, brain MRI)
Brain Death
- BAEP — absent all waves, or only Wave I present (cochlear/peripheral response without brainstem conduction)
- Presence of Wave I with absent Waves II–V confirms absent brainstem function with intact cochlea
- BAEP is an ancillary test for brain death determination (not sufficient alone)
Prognostication Summary Table
| Clinical Scenario | EP Finding | Interpretation |
|---|---|---|
| Post-cardiac arrest | Bilateral absent N20 (SSEP) | Poor prognosis (FPR <1%); most reliable single prognostic tool |
| Post-cardiac arrest | N20 present bilaterally | Does not guarantee good outcome (continue multimodal assessment) |
| Brain death | BAEP: absent all waves or only Wave I | Supports brain death (ancillary test) |
| Coma (traumatic) | Bilateral absent N20 (SSEP) | Poor prognosis but less reliable than in anoxic injury |
Clinical Pearl
SSEP-based prognostication after cardiac arrest should be performed at 24–72 hours and should not be the sole determinant. The 2023 AAN/AHA guidelines recommend a multimodal approach combining clinical exam, SSEP, EEG (absence of reactivity, status epilepticus), serum NSE levels, and brain imaging for neuroprognostication.
Quick Reference
Evoked Potentials at a Glance
| Question | Answer |
|---|---|
| Most sensitive EP for MS? | VEP (85–90% sensitivity) |
| Key VEP waveform? | P100 — prolonged latency = demyelination (optic neuritis) |
| Most robust BAEP wave? | Wave V (inferior colliculus) |
| BAEP in acoustic neuroma? | Prolonged I–III interpeak latency |
| BAEP in brain death? | Absent all waves or only Wave I present |
| Best prognostic test post-cardiac arrest? | SSEP: bilateral absent N20 at 24–72 hrs = poor outcome (FPR <1%) |
| SSEP pathway? | Dorsal columns → medial lemniscus → thalamus → sensory cortex |
| MEP pathway? | Motor cortex → corticospinal tract → anterior horn cell → muscle |
| Intraoperative alarm criteria? | 50% amplitude drop or 10% latency increase |
| SSEP alone can miss what? | Anterior cord ischemia (need MEP for corticospinal tract) |
| What does latency reflect? | Myelin integrity (prolonged = demyelination) |
| What does amplitude reflect? | Axonal integrity (reduced = axonal loss) |
BAEP Wave Generator Quick Table
| Wave | Generator | Quick Memory |
|---|---|---|
| I | Distal CN VIII | Ear (cochlear nerve) |
| II | Cochlear nucleus | Pontomedullary junction |
| III | Superior olivary complex | Lower pons |
| IV | Lateral lemniscus | Upper pons |
| V | Inferior colliculus | Midbrain (most robust) |
References
- Chiappa KH. Evoked Potentials in Clinical Medicine. 3rd ed. Lippincott-Raven; 1997.
- Aminoff MJ. Aminoff's Electrodiagnosis in Clinical Neurology. 6th ed. Elsevier; 2012.
- Nuwer MR. Fundamentals of evoked potentials and common clinical applications today. Electroencephalography and Clinical Neurophysiology. 1998;106(2):142–148.
- American Clinical Neurophysiology Society (ACNS). Guideline 9D: Guidelines on short-latency somatosensory evoked potentials.
- Wijdicks EFM et al. Practice parameter: Prediction of outcome in comatose survivors after cardiopulmonary resuscitation. Neurology. 2006;67(2):203–210.
- Continuum (AAN). Clinical Neurophysiology and Evoked Potentials review articles.