Curriculum/Pillar 6 · Advanced Electrophysiology

Intraoperative Neurophysiology

Module 27 of 36·28 min read

Learning objectives

1Integrate EMG, SSEP, and MEP modalities for IONM
2Apply alarm criteria for spinal cord and nerve-root monitoring
3Recognise anaesthetic and technical confounders intraoperatively

Intraoperative neurophysiological monitoring transplants the electrodiagnostic laboratory into the operating room, where its purpose inverts: not to diagnose established disease but to detect impending, reversible injury in real time, while the surgeon can still act. The discipline rests on continuous and provoked recording from motor and sensory pathways, a small set of validated alarm criteria, and an anaesthetic technique engineered to keep those signals interpretable.

Free-running and triggered EMG

Free-running EMG continuously monitors muscles innervated by nerves at risk. A surgically stressed nerve root or cranial nerve discharges neurotonicbursts — high-frequency, mechanically-evoked motor-unit trains that signal traction, compression, or thermal irritation. Brief, phasic bursts accompany blunt contact; sustained, sinusoidal "bomber" or "train" discharges warn of ongoing injurious stress and demand the surgeon pause. Free-running EMG is exquisitely sensitive but non-quantitative; it reports that a nerve is being irritated, not how much function remains.

Triggered EMG provides that quantitative dimension by deliberately stimulating tissue and measuring the response. Its canonical application is pedicle-screw stimulation: intact cortical bone insulates a correctly placed screw, so a high current is required to depolarise the adjacent nerve root, whereas a medial cortical breach exposes conductive tissue and the stimulation threshold falls. A response evoked at a low threshold (conventionally <~8–10 mA, with values <6 mA highly concerning) suggests the screw has violated the pedicle wall and lies dangerously close to the root. Triggered EMG also maps cranial nerves and identifies functional neural tissue within a tumour bed.

Somatosensory evoked potentials

Somatosensory evoked potentials (SSEPs) interrogate the large-fibre dorsal-column–medial-lemniscus pathway. Repetitive stimulation of a peripheral mixed nerve (median, ulnar, posterior tibial) evokes a volley ascending through dorsal columns, brainstem, thalamus and primary somatosensory cortex, averaged over hundreds of trials to extract the cortical response from background noise. SSEPs continuously survey the integrity of the posterior cord and the somatosensory projection during spinal deformity correction, aneurysm clipping, and posterior fossa surgery. Their cardinal limitation: they monitor only the dorsal sensory pathway and can remain normal while the anterior cord — and the corticospinal tracts that carry motor function — is rendered ischaemic, the historical basis for monitor-negative postoperative paraplegia.

Alarm criteria

The accepted SSEP warning threshold is a ~50% reduction in amplitude or a ~10% increase in latency from a stable baseline. For transcranial motor evoked potentials several criteria coexist: complete loss of the response, a marked amplitude reduction (thresholds of ~50–80% are used depending on protocol), or a sustained rise in the stimulation threshold required to elicit a response. These are alarm thresholds — triggers for immediate communication and intervention (check anaesthetic depth and blood pressure, reverse the surgical manoeuvre), not verdicts of permanent injury.

~50%

SSEP amplitude drop alarm

~10%

SSEP latency increase alarm

<~8–10 mA

Pedicle-screw breach threshold

~50–80% ↓

tcMEP warning (amplitude)

Motor evoked potentials

Transcranial electrical motor evoked potentials (tcMEPs) close the gap SSEPs leave open by monitoring the corticospinal tract directly. A brief high-voltage pulse train applied over the motor cortex activates corticospinal axons, and the descending volley is recorded either as compound muscle action potentials in target limb muscles (myogenic MEPs) or as a travelling D-wave from an epidural electrode. Because the anterior spinal artery supplies the corticospinal tracts, MEPs are the most sensitive monitor of anterior cord ischaemia. Their myogenic form is, however, profoundly sensitive to anaesthesia and to neuromuscular blockade, and their high-voltage stimulus produces patient movement — requiring coordination with the surgeon and relative contraindication in some intracranial procedures.

Cranial-nerve and nerve-root monitoring

In skull-base and posterior-fossa surgery, individual cranial nerves are monitored by free-running EMG from their target muscles (orbicularis oculi and oris for the facial nerve; tongue, palate, sternocleidomastoid for the lower cranial nerves) and mapped with a handheld stimulating probe to identify and preserve them within distorted anatomy. Facial-nerve monitoring during vestibular-schwannoma resection is the archetypal example, where a preserved, low-threshold response at the brainstem predicts postoperative facial function. In spinal surgery, nerve-root monitoring combines free-running EMG for irritation with triggered pedicle-screw testing, complementing the tract-level information from SSEPs and MEPs.

Clinical Pearl

A multimodal approach is the standard of care precisely because each modality monitors a different pathway. SSEPs watch the dorsal columns, MEPs the corticospinal tracts, free-running EMG the irritation of roots and cranial nerves, triggered EMG the proximity of hardware to neural tissue. The combination covers the territory no single modality can — the classic teaching being that intact SSEPs do not guarantee an intact motor system, which is the entire reason MEPs were added to spinal deformity monitoring.

Common Pitfall

Anaesthesia is the silent confounder. Volatile halogenated agents and nitrous oxide suppress the synaptically dependent components of MEPs and SSEPs in a dose-dependent fashion, so total intravenous anaesthesia (TIVA) — propofol with an opioid — is strongly preferred when MEPs are recorded. Neuromuscular blockade abolishes all EMG and myogenic MEP responses entirely. Before calling a loss of signal pathological, confirm that the anaesthetic has not deepened, that an inhalational agent or a relaxant bolus has not just been given, and that blood pressure has not fallen — anaesthetic and systemic changes mimic surgical injury and cause far more false alarms than true ones.

Key points

Free-running EMG detects neurotonic discharges that warn of nerve irritation; triggered EMG quantifies risk — a pedicle-screw threshold <~8–10 mA suggests medial cortical breach.
SSEPs monitor the dorsal-column pathway; alarm criteria are a ~50% amplitude drop or ~10% latency increase, but they can miss anterior-cord (motor) ischaemia.
tcMEPs monitor the corticospinal tract via transcranial electrical stimulation and are the most sensitive measure of anterior-cord ischaemia; criteria include signal loss, marked amplitude reduction, or threshold elevation.
Cranial-nerve and nerve-root monitoring combine free-running EMG for irritation with stimulated mapping to identify and preserve neural structures.
Volatile agents and nitrous oxide suppress MEPs/SSEPs (favour TIVA) and neuromuscular blockade abolishes EMG — always exclude anaesthetic and systemic causes before declaring a true alarm.