Central and Peripheral Motor Conduction Studies by Single-Pulse Magnetic Stimulation

Abstract

Single-pulse magnetic stimulation is the simplest type of transcranial magnetic stimulation (TMS). Muscle action potentials induced by applying TMS over the primary motor cortex are recorded with surface electromyography electrodes, and they are called motor-evoked potentials (MEPs). The amplitude and latency of MEPs are used for various analyses in clinical practice and research. The most commonly used parameter is the central motor conduction time (CMCT), which is measured using motor cortical and spinal nerve stimulation. In addition, stimulation at the foramen magnum or the conus medullaris can be combined with conventional CMCT measurements to evaluate various conduction parameters in the corticospinal tract more precisely, including the cortical–brainstem conduction time, brainstem–root conduction time, cortical–conus motor conduction time, and cauda equina conduction time. The cortical silent period is also a useful parameter for evaluating cortical excitability. Single-pulse magnetic stimulation is further used to analyze not only the central nervous system but also the peripheral nervous system, such as for detecting lesions in the proximal parts of peripheral nerves. In this review article we introduce four types of single-pulse magnetic stimulation—of the motor cortex, spinal nerve, foramen magnum, and conus medullaris—that are useful for the diagnosis, elucidation of pathophysiology, and evaluation of clinical conditions and therapeutic effects. Single-pulse magnetic stimulation is a clinically useful technique that all neurologists should learn.

Graphical Abstract

MAGNETIC STIMULATION

Magnetic stimulation was developed by Barker and coworkers in 1985, and it enables the human central or peripheral nervous system to be activated painlessly and noninvasively.¹,²,³,⁴ The magnetic stimulator employed in this technique includes a large capacitor that can store a large electrical charge. Discharging this large stored charge instantaneously into a coil induces a changing magnetic field around the coil in accordance with Faraday’s law. The resultant changing magnetic field induces electric currents in the body called induced currents or eddy currents. The induced currents flow in the opposite direction to the currents in the stimulation coil, and the principle of magnetic stimulation is to electrically activate neurons or nerves using these induced currents. Because the magnetic field can pass unimpeded through the bones (e.g., the skull), the induced currents can be made sufficient large to activate the human brain, which is electrically protected by the scalp and skull; this technique is called transcranial magnetic stimulation (TMS).

TMS over the primary motor cortex (M1) can activate cortical motoneurons (MNs) in M1; that is, the corticospinal tract (CST) neurons (CSTNs). Impulses (also called descending volleys) travel down the CSTs to fire spinal MNs via a single synapse on spinal MNs. These impulses then travel along the peripheral nerves to the muscles to finally activate the muscles via the neuromuscular junctions. The muscle activities elicited by TMS, called motor-evoked potentials (MEPs), are recorded using surface electromyography (EMG) electrodes. TMS over M1 is believed to activate CSTNs indirectly through activation of cortical interneurons connecting to CSTNs via a few synapses. Therefore, MEPs elicited by TMS reflect the excitability within M1 and are often used as a biomarker of M1 excitability.

MEPs are now widely used in clinical practice and in human movement research. Other single-pulse human brain stimulation methods include transcranial high-voltage electrical stimulation (TES)⁵,⁶,⁷,⁸,⁹,¹⁰,¹¹ and direct electrical stimulation of M1 following exposure by craniotomy.¹²,¹³ However, the pain resulting from current flow through the scalp impedes the wide use of TES, and the neurosurgical craniotomy required for direct electrical stimulation also prevents its routine use in clinical examinations.

Approximately 40 years have passed since the development of magnetic stimulation, during which diverse stimulation methods such as paired-pulse TMS and repetitive TMS (rTMS) have been developed²,³,⁴ in addition to single-pulse magnetic stimulation. However, clinicians (including neurologists) generally do not have a sufficient understanding of the clinical utility of magnetic stimulation. This review article therefore focuses on single-pulse magnetic stimulation, which is useful for the diagnosis, elucidation of pathophysiology, and evaluations of clinical conditions and therapeutic effects. There are several important neurophysiological parameters of MEPs, among which the central motor conduction time (CMCT) obtained by single-pulse magnetic stimulation is most commonly used as a clinical parameter. CMCT measurements can be used to objectively determine the presence or absence of CST lesions as well as the location of any lesion that is present. Herein we introduce several neurophysiological parameters, including CMCT.

Magnetic stimulator

MEPs should be elicited by a monophasic magnetic stimulator when making CMCT recordings.²,³,⁴ Such a stimulator can unidirectionally activate the central and peripheral nervous systems. In all types of single-pulse magnetic stimulation, the direction of the induced current has important effects on the efficient activation of neurons or nerves. It should be noted that a biphasic stimulator is not suitable for recording MEPs because the induced currents flow bidirectionally. Instead, biphasic stimulators are mainly used for rTMS to induce neuroplasticity in humans.

Magnetic coils

Four types of magnetic coils are commonly used in examinations (Fig. 1). A round coil is the simplest and most widely used type.¹,²,³,⁴ It is designed so that the induced currents are maximal underneath the edge of the coil, which is used for cortical and spinal nerve stimulation. A figure-of-eight coil (also called 8-shaped coil) is designed so that the induced currents are maximal just below where the two constituent round coil overlap. This is used for cortical focal stimulation, since its spatial resolution is approximately 1 cm.¹⁴ A double-cone coil combines two large round coils at an obtuse angle, allowing stimulation of deep structures beneath the site where the two coils overlap.¹⁵,¹⁶ This coil is used for stimulation of the leg motor area,¹⁵,¹⁶ foramen magnum,¹⁷,¹⁸,¹⁹,²⁰ and cerebellum.²¹,²²,²³ A magnetic augmented translumbosacral stimulation (MATS) coil is a large, round coil with a diameter of 20 cm, which induces the maximum currents just under the coil edge. This is a special coil for stimulating the lumbosacral spinal nerves and conus medullaris.²⁴,²⁵

Fig. 1. Different types of magnetic coils. MATS, magnetic augmented translumbosacral stimulation.

Types of magnetic stimulation

Single-pulse magnetic stimulation can be applied to the motor cortex, spinal nerve, foramen magnum, and conus medullaris.³,⁴ There are also various types of paired-pulse magnetic stimulation, including paired-pulse TMS over M1,²⁶,²⁷,²⁸ interhemispheric TMS,²⁹,³⁰,³¹ and cerebellar stimulation.²¹,²²,²³ For neuroplasticity induction, rTMS including conventional rTMS,³²,³³,³⁴ theta burst stimulation,³⁵,³⁶,³⁷ paired associative stimulation,³⁷,³⁸,³⁹ and quadripulse stimulation⁴⁰,⁴¹,⁴² can be used. Herein we focus on single-pulse magnetic stimulation.

Motor cortical stimulation

The CMCT is the most commonly used parameter in motor conduction studies, and it can be measured by cortical stimulation and spinal nerve stimulation. The CMCT mainly reflects CST conduction.³,⁴ For cortical stimulation, M1 contralateral to the recorded muscle should be activated. When activating the hand motor area with a round coil, its center is placed over the Cz location in the 10–20 international system, and the coil edges are located over the hand M1. When using a figure-of-eight coil, the center of the coil should be placed over the hand M1 (at C3 or C4). The stimulating current is usually directed forward and around 45° medially to ensure that the induced currents are perpendicular to the central sulcus. A round or double-cone coil is used to activate the leg motor area, with the latter normally used since it readily activates the deeply sited leg M1. The edge of a round coil or the center of a double-cone coil is placed over Cz, and the induced currents should be directed from Cz to the contralateral side to the recorded muscle. With these settings the coil activation site is over the leg M1, and the induced currents must be directed parallel to the CST axons of the descending tracts from the leg M1 (Fig. 2). Because magnetic fields attenuate with the square of the distance, the coil should be in close contact with the scalp.

Fig. 2. Coil position and induced-current direction when recording motor-evoked potentials on the right side. The left and right columns show the recordings from hand and leg muscles, respectively. For hand-muscle recordings, a round coil is placed over the Cz location and the coil edge is located over the hand motor area, or the center of a figure-of-eight coil is placed over the hand motor area. For leg-muscle recordings, the posterior edge of a round coil is placed over Cz (leg motor area), or the center of a double-cone coil is placed over Cz. Solid arrow, coil-current direction; dashed arrow, induced-current direction.

When the recorded muscle is volitionally contracted during cortical stimulation, the MEP amplitude is higher and its latency is shorter in the so-called active condition relative to the so-called resting condition when the muscle is relaxed. There are two main reasons why the amplitude and latency of the MEP differ between the active and resting conditions: 1) multiple descending volleys are induced by a single cortical stimulation in the CSTs and 2) spatial and temporal summations of excitatory postsynaptic potentials (EPSPs) at the synapses between descending tracts and spinal MNs.

Multiple descending volleys

Electrical stimulation of peripheral nerves generates single antegrade and retrograde impulses in the axons of peripheral nerve; the former elicits muscle contractions via the neuromuscular junctions and the latter elicits F-waves. In contrast, TMS generates multiple descending impulses within the CSTs,³,⁴ called multiple descending volleys. These multiple impulses are generated at intervals of approximately 1.5 ms and travel along the descending tracts including the CSTs.⁴³ TMS generally does not activate CSTNs directly, but instead activates them via a few synapses within M1 by directly activating motor cortical interneurons connected to CSTNs. The indirect activation of CSTNs means that the impulses are called I-waves (for Indirect), and they are named I1-wave, I2-wave, I3-wave, etc. Which type of I-wave is generated depends on the induced-current direction; for example, an I1-wave is more likely to be produced when the induced currents flow forward over the hand M1, while an I3-wave is more likely when the induced currents flow backward. The I1-wave preferential activation method is normally used for neurophysiological examinations.

TES directly activates and fires CSTNs in M1. Therefore, an impulse with a latency of approximately 1.5 ms earlier than the I1-wave is generated first, which is called the D-wave (for Direct). When the stimulus intensity is increased, I-waves appear after D-waves. Note that even during magnetic stimulation, a D-wave can be generated by TMS with a high stimulus intensity.

EPSP summations at the spinal MNs

The MEP latency is approximately 3 ms shorter in the active condition than in the resting condition.³,⁴ This can be explained by the spatial and temporal EPSP summations at the synapses of spinal MNs: for firing spinal MNs, spatial and temporal EPSPs summations are required, where EPSPs are elicited by the multiple descending volleys. The membrane potentials of spinal MNs are low in the resting condition, and when an I1-wave reaches them the membrane potentials depolarized by the I1-wave still do not reach the threshold required for the spinal MNs to fire. Therefore, while waiting for the arrival of later I-waves such as an I3-wave, spinal MNs can fire for the first time due to the EPSP temporal summations (Fig. 3, upper panels). In contrast, when an I1-wave reaches in the active condition, although some spinal MNs are now firing or are in the refractory period after being fired, many of the remaining spinal MNs are in a depolarized state below the threshold because the baseline membrane potentials are generally higher in the active condition than in the resting condition. The arrival of an I1-wave at the spinal MNs in the active condition can result in some spinal MNs firing (Fig. 3, lower panels). Some more spinal MNs will then fire when an I2-wave or I3-wave arrives. This results in the onset latency of the potential (the MEP) generated by all of the spinal MNs being shortened by approximately 3 ms compared with that in the resting condition. Therefore, spinal MNs will usually fire when an I1-wave arrives in the active condition and when an I3-wave arrives in the resting condition, corresponding to the latency difference of approximately 3 ms; that is, I3-wave latency–I1-wave latency≅3 ms.

Fig. 3. Temporal excitatory postsynaptic potential (EPSP) summations at the spinal motoneurons (MNs). Membrane potentials at the spinal MNs in the resting condition (upper panels) and active condition (lower panels). EPSPs are generated by multiple successive descending volleys consisting of the indirect I-waves. In the resting condition, a few spinal MNs are able to fire when the I3-wave arrives, with most spinal MNs failing to fire. Therefore, the MEP latency depends on the I3-wave, and the MEP is small. In the active condition, most spinal MNs are able to fire when I-waves arrive, and so the MEP latency depends on the I1-wave, and the MEP is large.

The MEP is larger in the active condition than in the resting condition. In the resting condition, the membrane potential is low and most spinal MNs fail to fire when the EPSP is evoked by I-waves, while a few spinal MNs can fire when applying high-intensity stimulation (Fig. 3, upper panels). In contrast, because most spinal MNs are in a depolarized state below the activation threshold before stimulation in the active condition the arrival of I-waves leads to firing of a considerable number of spinal MNs, and many of them will fire when the I3-wave arrives (Fig. 3, lower panels).

The MEP latency is not affected by the strength of muscle contraction because firing occurs upon the arrival of an I1-wave in both weak and strong muscle contraction conditions. In general, the MEP latency is the shortest even with weak muscle contraction, such as 5%–10% of the maximum muscle contraction.⁴⁴ In contrast, the MEP amplitude is affected by the strength of muscle contraction. A greater contraction results in a larger MEP, probably because more spinal MNs are in a depolarized state near to the activation threshold in the strongly active condition. However, strong muscle contraction makes it difficult to detect the onset latency of MEPs due to the presence of large baseline fluctuations. Therefore, when measuring the CMCT it is recommended to record MEPs under a condition of weak muscle contraction. Only when an MEP is not elicited during weak muscle contraction should strong, phasic muscle contraction be tried just to obtain MEPs. Compared with tonic muscle contraction, phasic muscle contraction produces a large number of descending volleys arriving at spinal MNs simultaneously during the contraction, which will result in the depolarization of many spinal MNs that are near to the activation threshold. Under this condition, when the descending volleys induced by TMS reach the spinal MNs, many of them can fire, markedly increasing the MEPs.⁴⁵

In summary, the MEP latency depends on which I-wave fires spinal MNs, and the MEP amplitude depends on the number of spinal MNs fired. The MEP amplitude is also affected by the strength of muscle contraction.

EXCEPTIONAL CASES

Case 1: ipsilateral CST projections

This subject was a young healthy male volunteer. Applying TMS over the left-hand M1 using a figure-of-eight coil did not elicit clear MEPs from the right-hand muscle, but MEPs from the left-hand muscle were clearly recorded (Fig. 4, left column). Similarly, applying TMS over the right M1 resulted in MEPs from the right-hand muscle being clearly recorded (Fig. 4, right column). Small MEPs followed by a short silent period (SP) were evoked in the left-hand muscle. These MEP latencies were approximately 20 ms, indicating the involvement of direct monosynaptic ipsilateral projections from the CSTNs to the cervical spinal MNs. To exclude current spread to the contralateral motor cortices, we applied TMS over Cz at the same intensity as that used to stimulate the hand-motor cortical area, which did not elicit MEPs (Fig. 4, middle column). In summary, this TMS study demonstrated ipsilateral motor cortical projections in a healthy subject, probably reflecting incomplete pyramidal decussation.

Fig. 4. Case 1: ipsilateral corticospinal tract projections. Exceptional transcranial magnetic stimulation (TMS) findings in a healthy subject. The upper and lower rows show MEPs from the right and left first dorsal interosseous (FDI), respectively. The left, middle, and right columns show MEPs for TMS applied over the left-hand primary motor cortex (M1), over Cz, and over the right-hand M1, respectively. MEPs from the left and right FDI were elicited by TMS over the left-hand and right-hand M1, respectively. No MEPs were elicited by applying TMS over Cz. MEP, motor-evoked potential.

Case 2: bilateral CST projections from the unilateral M1

This patient was a female who had right-hand paresis due to a birth injury but no leg motor paresis. Brain magnetic resonance imaging (MRI) revealed a left frontal cerebral infarction involving the hand motor area (Fig. 5, left panel). She was able to move her left hand smoothly associated with synchronous right-hand movements, whereas her right hand moved clumsily in associated with synchronous left-hand movements. This phenomenon is called “mirror movements.” Applying TMS over the left-hand M1 using a figure-of-eight coil did not elicit MEPs from the muscles of either hand. In contrast, applying TMS over the right-hand M1 resulted in MEPs being recorded from both hand muscles (Fig. 5, right panel). The MEP was smaller in the right hand than in the left hand. The MEP latencies were approximately 20 ms, indicating the involvement of direct monosynaptic projections from the right M1 to the bilateral cervical spinal MNs. In summary, this TMS study demonstrated that bilateral upper limb muscles were innervated by the right M1 via the CST projections in this patient.

Fig. 5. Case 2: bilateral CST projections from the unilateral M1. TMS findings in the patient with mirror movements. Brain magnetic resonance imaging shows a left frontal cerebral infarction (left panel). MEPs were not elicited from either of the hand muscles when TMS was applied over the left-hand M1. In contrast, MEPs were elicited in both hand muscles when TMS was applied over the right-hand M1. The arrows indicate the stimulus point (right panel). CST, corticospinal tract; FDI, first dorsal interosseous; M1, primary motor cortex; MEP, motor-evoked potential; TMS, transcranial magnetic stimulation.

Case 3: motor cortex isolation

A female showed left hemiplegia with no voluntary movements on the left side. She had cerebral infarctions in areas perfused by the right anterior cerebral artery, including the medial frontal cortex, corpus callosum, leg motor area, and parietal cortex, while the hand motor area was preserved. Applying TMS over the right M1 elicited normal MEPs in the left-hand muscle, with the loss of interhemispheric connections from the left M1. The hemiplegia of this patient was not caused by CST involvement; rather, it was caused by disconnection of the intact M1 from the structures required for initiating movements.⁴⁶

Spinal nerve stimulation

The spinal nerve stimulation technique is also called motor root stimulation. A round coil is placed over the neck when recording MEPs in the upper limbs, whereas a round or MATS coil is placed over the lumbosacral region when recording MEPs in the lower limbs.⁴⁷,⁴⁸ Because the lumbosacral spinal nerves are deep, it is easier to elicit MEPs when using an MATS coil for lumbosacral stimulation.²⁴,⁴⁸ No currents are induced in the bone due to its very low electric conductivity, and so will be concentrated in the intervertebral foramina, making it possible to activate spinal nerves at the intervertebral foramina.⁴⁹ The MEP latency is constant during spinal nerve stimulation regardless of the coil position, since the induced currents are concentrated in the intervertebral foramina and preferentially activate large nerve fibers of the spinal nerves. Spinal nerve stimulation should be performed in the resting condition because it is a type of peripheral nerve stimulation.

To elicit large MEPs using a round coil, the upper edge of the coil should be placed over the target spinal nerves; namely, the C7 spinous process for hand muscles and the L5 spinous process for leg muscles, which should induce a distal-to-proximal current flow. When using a MATS coil to record MEPs from the lower limbs, the coil should be placed over the lumbosacral regions on the contralateral side of the recorded muscle, and the coil edge should be placed along the spinal nerves. Eliciting large MEPs requires the induction of a proximal-to-distal current flow, which is opposite to the direction for a round coil.²⁴,⁴⁸

Stimulation at the foramen magnum

The technique for stimulating CSTs at the foramen magnum is also called brainstem stimulation. The center of a double-cone coil should be placed over the external occipital protuberance (i.e., the inion), and induced currents should flow upward. Because the induced currents are concentrated in the foramen magnum, it is possible to activate CSTs at the pyramidal decussation level.¹⁷,¹⁸,¹⁹,²⁰ The stimulus threshold is higher for stimulation at the foramen magnum than for other stimulation methods. In some subjects, placing the center of a double-cone coil slightly toward the recording side relative to the inion is better than placing it directly over the inion.⁵⁰

Stimulation at the foramen magnum generates a single descending volley because this involves the axonal stimulation of CSTs.¹⁰,¹⁷ Spinal MNs can fire due to a single EPSP produced by a single impulse. The lack of temporal EPSP summations means that the MEP latency does not differ between the active and resting conditions. On the other hand, the MEP amplitude increases in the active condition, which is due to most spinal MNs being in a depolarized state below the activation threshold during muscle contraction, and many spinal MNs can fire with a single EPSP. The constant latency of the MEP in response to stimulation at the foramen magnum is probably due to the induced currents being concentrated in the foramen magnum. MEPs often cannot be elicited in patients with severe CST damage. In such cases, double-pulse stimulation at the foramen magnum might elicit clear MEPs by artificially inducing the temporal summations of EPSPs in the spinal MNs.⁵¹,⁵²

Stimulation at the foramen magnum can also be used in research into changes in spinal motoneuronal excitability.²⁰ The elicitation of large MEPs by such stimulation could indicate high spinal cord excitability.

Conus stimulation

A MATS coil should be placed over the lumbosacral regions on the contralateral side of the recorded muscle when stimulating the conus medullaris. The coil edge should be placed over the L1 spinous process and the induced currents should flow upward. Because the induced currents are concentrated around the conus medullaris, it is possible to activate the root exit zone of the cauda equina from the conus medullaris in the spinal canal.²⁵,⁴⁸ The MEP latency during conus stimulation is constant, regardless of its amplitude. This is explained by the induced currents being concentrated around the conus medullaris and preferentially activating large nerve fibers of the cauda equina.⁵³,⁵⁴ Conus stimulation should be performed in the resting condition.

The four types of single-pulse magnetic stimulation in addition to conventional electrical stimulation of peripheral nerves are summarized in Table 1.

Table 1. Single-pulse magnetic stimulation methods.

Stimulation site	Volleys in axons	Synapses	Active condition
			MEP amplitude	MEP latency
Peripheral nerves	1	NMJ	No change	No change
Spinal nerves	1	NMJ	No change	No change
Conus medullaris	1	NMJ	No change	No change
Foramen magnum level	1	NMJ, spinal MNs	Increased	No change
Cortex	3 or 4	NMJ, spinal MNs, cortical MNs	Increased	Shortened

MEP, motor-evoked potential; MN, motoneuron; NMJ, neuromuscular junction.

NEUROPHYSIOLOGICAL PARAMETERS

Motor threshold

The motor threshold in cortical stimulation reflects the excitability of the entire descending pathways from M1 to the recorded muscle. This parameter is often used as a reference value for setting a stimulus intensity in experiments using single-pulse TMS (e.g., the cortical silent period [CSP]), paired-pulse TMS (e.g., short-interval intracortical inhibition and intracortical facilitation), and rTMS. The minimum stimulus intensity at which an MEP is elicited in the resting condition is called the resting motor threshold (RMT), and that in the active condition is called the active motor threshold (AMT). The RMT and AMT are defined as the minimum stimulus intensities required to elicit MEPs of ≥50 µV and ≥200 µV, respectively, with a probability of >0.5.³,⁴

A round or figure-of-eight coil is used to record MEPs from the upper limbs, whereas a round or double-cone coil is used for the lower limbs. The usual coil positions and induced-current directions are shown in Fig. 2, and the coil should be positioned at the site with the lowest threshold (i.e., hotspot).

An abnormal high threshold usually indicates the presence of CST damage, but this is not definitive because the motor threshold varies widely among individuals. Note that the motor threshold is increased by sodium-channel blockers such as carbamazepine, phenytoin, and lamotrigine, and is decreased by the N-methyl-D-aspartate antagonist ketamine.⁵⁵

Central motor conduction time

The CMCT is measured when evaluating CST conduction, with usually both cortical and spinal nerves being stimulated.³,⁴ This parameter is obtained by subtracting the latency MEP to spinal nerve stimulation (i.e., spinal latency or root latency) from the latency obtained by cortical stimulation (i.e., cortical latency): CMCT=cortical latency–spinal latency. The CMCT can also be calculated by subtracting the peripheral motor conduction time obtained by F-waves from the cortical latency: CMCT=cortical latency − (F+M − 1)/2 (F: F-wave latency, M: M-wave latency). The advantage of this measurement using F-waves is that the CMCT does not include any peripheral nerve components. However, F-waves themselves are often not elicited in patients with peripheral neuropathy.

Cortical stimulation should be performed in the active condition; that is, during weak muscle contraction. Determining the MEP latency usually requires four to eight reproducible MEP waveforms to be recorded, with their superimposition used to measure the latency. When it is difficult to elicit MEPs in patients with CST involvement or certain disorders, TMS can be repeated 10 to 20 times and the onset latency can then be measured using the averaged MEP waveforms. Spinal nerve stimulation should be performed in the resting condition, and there is little variation in the MEP waveforms. Therefore, the reproducibility can be easily confirmed usually after four measurements.

Causes of CMCT prolongation

The CMCT can be prolonged by demyelination of the CSTs, as in multiple sclerosis,⁵⁶ or by axonal degeneration, as in amyotrophic lateral sclerosis.⁵⁷,⁵⁸ In general, the sensitivity of MEP abnormalities is higher in demyelination than in axonal degeneration.³,⁴ This is explained as follows: The main mechanism underlying CMCT prolongation in axonal degeneration is impairment of EPSP summations at the spinal MNs. Because the CMCT would be prolonged by only a few milliseconds when EPSP summation fails, it is not easily detected, resulting in low sensitivity in axonal degeneration. However, severe damage to the CSTs even in axonal degeneration could result in extreme prolongation of the CMCT since the stimulus threshold of the CSTs increases and MEPs are induced by the activation of other slowly conducting descending tracts, such as rubrospinal, vestibulospinal, reticulospinal, and tectospinal tracts.

It should be noted that CMCT prolongation does not necessarily indicate CST damage. Patients with dementia or a functional neurological disorder often have difficulty in voluntary muscle contraction, which leads to recordings being made in the resting condition. In such cases the cortical latency would be prolonged by a few milliseconds. Another reason for CMCT prolongation is conduction delay at the most-proximal parts of peripheral nerves in patients with severe demyelinating neuropathy, because spinal nerve stimulation activates spinal nerves at the intervertebral foramina.¹⁸,⁴⁸ Because the CMCT includes the short conduction time of spinal nerves from spinal MNs to the intervertebral foramina in the spinal canal, CMCT could be prolonged in peripheral neuropathy. The causes of CMCT prolongation are summarized in Table 2.

Table 2. Causes of prolongation of the central motor conduction time.

Cause		Example disorders
1	Demyelination of CSTs	Multiple sclerosis
2	Axonal degeneration of CSTs (impaired EPSP summations at spinal MNs)	Amyotrophic lateral sclerosis
3	Increased stimulus threshold of CSTs (activation of other descending tracts)	Amyotrophic lateral sclerosis
4	Difficulty of voluntary muscle contraction	Dementia, functional neurological disorder
5	Conduction delay of spinal nerves	Demyelinating neuropathy

CST, corticospinal tract; EPSP, excitatory postsynaptic potential; MN, motoneuron.

Care must be taken to account for how a subject’s height and age influence the CMCT, in particular regarding the lower limbs. Therefore, to identify CMCT abnormalities, measured CMCT values should be compared with values from body-height-matched and age-matched healthy subjects.

Case 4: abnormal TMS findings in a patient with normal MRI findings

This patient was a male with spinal cord infarction who suddenly presented with right-leg palsy. The findings of neurological examinations were compatible with right-sided Brown-Séquard syndrome, suggesting the involvement of the right half of the spinal cord. However, spinal MRI revealed no abnormal findings (Fig. 6, right panel, day 2). A TMS examination revealed that the CMCT was normal in both hand muscles and the left-leg muscle, but prolonged in the right-leg muscle (Fig. 6, left panels). A TMS study suggested the involvement of CSTs at the right thoracic-lumbar spinal cord levels because CST conduction from M1 to right lumbar region was involved but conduction from M1 was preserved to the right cervical regions and the left lumbar region. Two weeks later his right-leg palsy had improved and T2-weighted MRI revealed hyperintense lesions in the right posterior part of the spinal cord, indicating posterior spinal artery syndrome (Fig. 6, right panel, day 14).

Fig. 6. Case 4: abnormal TMS findings in a patient with normal MRI findings. Central motor conduction time (CMCT) findings in a patient with spinal cord infarction who suddenly presented with right-leg palsy. On day 2, spinal MRI reveals no abnormal findings (right panel, day 2). A TMS study showed prolonged CMCT only in the right-leg muscle, suggesting a right thoracic-lumbar lesion (left panels). On day 14, spinal MRI showed spinal cord lesions supplied by the right posterior spinal artery (right panel, day 14). FDI, first dorsal interosseous; MRI, magnetic resonance imaging; TA, tibialis anterior; TMS, transcranial magnetic stimulation.

In summary, this case shows that Brown-Séquard syndrome can occur transiently, even in posterior spinal artery syndrome, as reported previously.⁵⁹ Furthermore, the present study demonstrated that TMS can be used to localize lesions even when MRI does not reveal any lesions in the acute phase.

Cortical–brainstem and brainstem–root conduction times

Prolongation of the CMCT is suggestive of CST involvement, but it is difficult to localize a lesion within the CSTs. Stimulation at the foramen magnum can be used to measure the cortical–brainstem conduction time (C-BSTCT) and the brainstem–root conduction time (BST-RCT).¹⁷,¹⁸,¹⁹,²⁰ Measurements of these conduction times can be used to identify whether the conduction delay occurs at an intracranial site, an extracranial site, or both. The C-BSTCT is calculated by subtracting the brainstem latency from the cortical latency (i.e., C-BSTCT=cortical latency − brainstem latency) and the BST-RCT is calculated by subtracting the spinal latency from the brainstem latency (i.e., BST-RCT=brainstem latency − spinal latency).

The methods used to obtain the C-BSTCT and BST-RCT are similar to those used for cortical stimulation. The stimulation at the foramen magnum should be done in the active condition, with usually four to eight reproducible MEP waveforms being required.

The clinical utility of stimulation at the foramen magnum has been demonstrated in several neurological disorders, including amyotrophic lateral sclerosis, cervical spondylosis, multiple sclerosis, stroke, and polyneuropathy.¹⁸,⁶⁰ The main advantages are the abilities to 1) detect a subclinical CST lesion, 2) demonstrate multiple CST lesions, 3) identify two disorders, and 4) reveal a CST lesion that is clinically masked by peripheral nerve involvement.¹⁸ The conduction characteristics of slowly conducting descending tracts other than CSTs might be evaluable in patients with severe damage to the CSTs.⁶¹

Cortical–conus motor and cauda equina conduction times

The cauda equina comprises long peripheral nerves in the lumbosacral spinal canal, and so measurements of the CMCT from the lower limb muscles include the conduction time in the cauda equina. It is important to note that peripheral nerve conduction times are included in the CMCT of patients with peripheral neuropathy, especially demyelinating neuropathies. This challenge is usually surmounted by using conus stimulation, which yields the cortical–conus motor conduction time (CCCT) and the cauda equina conduction time (CECT)⁶²,⁶³; the CCCT reflects purely CST conduction and the CECT reflects peripheral nerve conduction. Knowledge of these two values make it possible to localize conduction delays. The CCCT is calculated by subtracting the conus latency from the cortical latency (i.e., CCCT=cortical latency − conus latency) and the CECT is calculated by subtracting the spinal latency from the conus latency (i.e., CECT=conus latency − spinal latency).

To determine the MEP latency, conus stimulation should be applied in the resting condition until reproducible MEPs are obtained, which is usually possible within four attempts.

The CCCT is less influenced by body height and age. The smaller effect of body height might be due to disproportionate growth in the length of the spinal cord and vertebral column.⁶² Meanwhile, the smaller effect of age on the CCCT might be explained by the protection of the central nervous system by bones, cerebrospinal fluid, and the blood–brain barrier.⁶³ In contrast, the CECT is strongly influenced by both body height and age. The CCCT can be used to accurately estimate CST conduction time, even in neurological disorders involving peripheral nerves.⁶⁴,⁶⁵

Case 5: CST involvement with peripheral neuropathy

This patient was a 60-year-old female with adult-onset Krabbe disease who presented with slowly progressive gait disturbance due to mild spastic paraplegia. A TMS study revealed severe prolongation of both spinal and conus latencies. The problem in this case was that the conventional CMCT did not accurately reflect the CST conduction because the CMCT can be prolonged by peripheral neuropathy. In contrast, conus stimulation demonstrated that both the CCCT and CECT were prolonged. This indicates that CST involvement in addition to peripheral neuropathy were present in this patient.⁶⁴

Cortical silent period

The CSP is used to evaluate motor cortical inhibition using single-pulse TMS. Applying TMS in the active condition transiently stops voluntary muscle contraction immediately after MEPs.⁶⁶ The time during which muscle contraction does not occur is called the CSP. The former part of the CSP (from 0 ms to 60 ms) is believed to reflect spinal cord inhibition, and the latter part reflects motor cortical inhibition.³,⁴ Fig. 7 shows the SPs induced by stimulation over the motor cortex (CSP) and at foramen magnum (SP). The MEP amplitude increased and the SP period was prolonged by strong cortical stimulation (150 ms at a stimulation intensity of 70% of the maximum output), but the SP for stimulation at the foramen magnum saturated at around 60 ms and was not prolonged by stimulation at a higher intensity. This supports the hypothesis that the former part of the CSP is generated by spinal mechanisms and the latter by cortical mechanisms.

Fig. 7. Silent periods (SPs) induced by TMS and by stimulation at the foramen magnum. Increasing the intensity of cortical stimulation increased the MEP amplitude and prolonged the SP. In stimulation at the foramen magnum, SP duration saturated at around 60 ms. MEP, motor-evoked potential; TMS, transcranial magnetic stimulation.

MEPs are usually recorded from a hand muscle when a round or figure-of-eight coil is placed over the motor hotspot. The coil position and induced-current direction are determined by minimizing the threshold (Fig. 2). The RMT and AMT are measured before recording the CSP. The stimulus intensity is often set at 120%–140% of the RMT or AMT, and typically 8 to 12 MEPs are recorded. The CSP is usually defined as the time from the onset of an MEP to when EMG activities reappear after they have been suppressed.

The duration of the CSP is typically 100 ms to 200 ms in healthy subjects, but there is a large interindividual variability. It is also strongly influenced by the stimulus intensity, with a stronger stimulus intensity resulting in a longer CSP, whereas the degree of muscle contraction does not affect the CSP duration. In a CSP examination, TMS is applied at a certain intensity while the subject maintains a constant active condition, such as 20% of maximal muscle contraction.

Note that the CSP is strongly affected by drugs. In particular, gamma-amino butyric acid (GABA) agonists prolong the CSP, such as the GABA_A receptor agonists lorazepam and zolpidem, and the GABA_B receptor agonist baclofen. Levodopa and dopamine receptor agonists can also increase the sensitivity of GABA receptors and prolong the CSP.⁵⁵ Shortening or lengthening of the CSP is observed in various neurological disorders. For example, the CSP is markedly shortened in patients with cortical myoclonus⁶⁷ and markedly lengthened in patients with Creutzfeldt-Jakob disease (CJD).⁶⁷,⁶⁸,⁶⁹

Case 6: an early neurophysiological biomarker of CJD

This patient was a 66-year-old female who developed rapidly progressive dementia. She did not show any myoclonic jerks. Electroencephalography showed only diffuse and continuous slow waves. Diffusion-weighted brain MRI revealed hyperintense lesions in the cerebral cortex, basal ganglia, and thalamus. A TMS study revealed that the CSP was markedly prolonged. She subsequently exhibited akinetic mutism and developed periodic myoclonus in her limbs. Electroencephalography showed periodic synchronous discharges. The final diagnosis was CJD. Based on this case and the previous literature, we propose that marked prolongation of the CSP could be an early neurophysiological biomarker of CJD.⁶⁸

PERIPHERAL NERVOUS SYSTEM

Spinal nerve stimulation

Spinal nerve stimulation is usually applied for CMCT measurements, as described above. In addition to central motor conduction studies, spinal nerve stimulation is used to evaluate the most-proximal parts of peripheral nerves. In such cases the MEPs elicited by spinal nerve stimulation are also called compound muscle action potentials (CMAPs), like in a nerve conduction study (NCS). To achieve supramaximal stimulation, a round coil is needed for the upper limb muscles and a MATS coil is used for the lower limb muscles.²⁴,⁴⁸,⁷⁰ Supramaximal stimulation should be confirmed using a conventional method in NCS. Supramaximal stimulation is not achievable in some patients with peripheral neuropathy, even when the stimulator is set to its maximum output. In such patients only the latencies are evaluated.

When using a round coil, induced currents should flow in a distal-to-proximal direction along the target spinal nerves. The optimal site for eliciting CMAPs (i.e., the hotspot) is likely to be around the C7 spinous process for the hand muscles. When using a MATS coil, induced currents should flow along the target spinal nerves in a proximal-to-distal direction, where the direction of the induced current is opposite that in the round coil. The hotspot is likely to be located around the L5 spinous process for the lower limb muscles. At the hotspot, the stimulus intensity should be gradually increased to maximize the CMAPs. This technique can be combined with a conventional NCS at the distal parts of limbs.

For the upper limb muscles, focal lesions between Erb’s point and the intervertebral foramina (i.e., brachial plexus or spinal nerves just distal to the intervertebral foramina) can be detected. For example, a conduction block in the upper or lower part of the brachial plexus could be detected by stimulating the spinal nerve in patients with neuralgic amyotrophy or tumor invasion.⁷⁰,⁷¹ For the lower limb muscles, focal lesions between the knee and the intervertebral foramina (i.e., sacral nerves, sacral plexus, or spinal nerves just distal to the intervertebral foramina) can also be detected. For example, a conduction block was detected in the sciatic nerve by applying spinal nerve stimulation in a patient with sacral nerve palsy.⁷² In two patients with neuralgic amyotrophy and sacral nerve palsy, conduction blocks were improved and CMAP amplitudes were normalized with intravenous immunoglobulin treatment.⁷⁰,⁷²

Case 7: pseudo muscular weakness (poor motor performance) in acute pure sensory neuropathy

This patient was a female who complained of acute muscular weakness in the extremities. She became bedridden and was unable to move her hands voluntarily. However, manual muscle tests showed that her muscle power was sufficient to resist the forces applied to her arms by the examiner. A conventional NCS and spinal nerve stimulation elicited normal-amplitude CMAPs with normal latencies, suggesting that the peripheral motor nerves were intact (Fig. 8, left panels). In contrast, no median nerve sensory-evoked potentials were evoked, suggesting severe damage to the sensory nervous systems (Fig. 8, lower-right panel, before treatment). Lumbar T1-weighted MRI showed gadolinium enhancement bilaterally in the posterior roots (Fig. 8, upper-right panel). Anti-GD1b ganglioside antibody was seropositive; GD1b has been shown to be localized in the posterior spinal roots.⁷³ After intravenous immunoglobulin treatment, the patient was able to move her extremities, and normal-amplitude sensory-evoked potentials were elicited (Fig. 8, lower-right panel, after treatment). In this case, neurophysiological examinations supported that her muscular weakness was due to a pure sensory neuropathy without any involvement of motor nerves (i.e., pseudo muscular weakness).

Fig. 8. Case 7: pseudo muscular weakness (poor motor performance) in acute pure sensory neuropathy. Neurophysiological and neuroradiological findings in a patient who developed acute muscular weakness in the extremities. A nerve conduction study and spinal nerve stimulation demonstrated normal findings, suggesting that the peripheral motor nerves were intact (left panels). In contrast, median nerve sensory-evoked potentials were not elicited (lower-right panel, before treatment), suggesting the presence of severe damage to the sensory nervous systems. Lumbar MRI showed gadolinium enhancement bilaterally in the posterior roots (yellow arrows) of the cauda equina (upper-right panel). After treatment with intravenous immunoglobulin, sensory-evoked potentials were elicited (lower-right panel, after treatment). ADM, abductor digiti minimi; FDI, first dorsal interosseous; MRI, magnetic resonance imaging; T1WI, T1 weighted image.

Conus stimulation

The CCCT and CECT are also used for assessing central motor conduction and peripheral nerve condition. The former application is described above. The CECT can sometimes provide critical information about proximal peripheral nerves.

The currents induced by a MATS coil should flow upward in the body. The hotspot is likely to be located at the L1 spinous process. At the hotspot, the stimulus intensity should be gradually increased until a stable CMAP is elicited.²⁵,⁴⁸ This technique can be combined with a conventional NCS and spinal nerve stimulation. The latency of CMAPs to conus stimulation is usually stable and constant, but the CMAP amplitude is often not supramaximal; it is important to remember this when analyzing the results.

The CECT is more frequently prolonged than the distal segment of peripheral nerves in patients with Guillain-Barré syndrome (GBS)⁷⁴ and chronic inflammatory demyelinating polyradiculoneuropathy (CIDP).⁷⁵,⁷⁶ CECT prolongation might be an early biomarker of the demyelinating forms of GBS and CIDP.⁷⁴,⁷⁵,⁷⁶ The CECT is also prolonged in other peripheral neuropathies, such as POEMS (polyneuropathy, organomegaly, endocrinopathy, monoclonal plasma cell disorder, skin changes) syndrome,⁷⁷ anti-myelin-associated glycoprotein polyneuropathy,⁴⁸ tumor invasion,²⁴ and postradiation lower motor neuron syndrome.⁷⁸

DIFFICULTIES OF RECORDING SESSIONS

Examination room

Hundreds of measurements are required to record evoked potentials such as sensory-evoked potentials, auditory brainstem responses, and visual-evoked potentials due to the waveforms being very small (on the order of microvolts) and often being buried in noise. In contrast, MEPs are barely affected by alternating-current noise from devices such as medical equipment and computers because the MEP waveforms are quite large (on the order of millivolts).²,³,⁴ However, sometimes the signals will be contaminated by very large alternating currents, making it preferable to perform TMS in an EMG examination room that is electrically shielded. In addition, the large shock artifact of magnetic stimulation makes it necessary to maintain a sufficient distance between magnetic stimulators and recording devices. In a small examination room the coil cable may need to be coiled up, which will itself generate a large magnetic field, and so a large examination room is desirable to allow the coil cable to be uncoiled. During examinations, the subject sits on a bed or chair or lies down on a reclining chair or bed and must be fully relaxed. Moreover, the excitability of the cerebral cortex is affected by even small sounds, and so a quiet environment is required, and the subject should have their eyes open to keep awake during examinations.

Target muscle selection

The first dorsal interosseous or abductor digiti minimi is often used in CMCT analyses of hand muscles. These muscles are used due to the ease of voluntary muscle contraction and the smaller influence of volume conduction from surrounding nontarget muscles.⁷⁹ The abductor pollicis brevis is relatively rarely used because of strong volume conduction effects on this muscle. When stimulating the cortex, brainstem, or spinal nerve, impulses travel down both the median and ulnar nerves. The resulting simultaneous activation of multiple nerves when applying these methods means that volume conduction should be considered more than in a conventional NCS. In particular, this volume conduction effect will markedly affect the results when the recorded muscle is atrophic.

The tibialis anterior (TA) and abductor hallucis (AH) are often used in studies of lower limb muscles. The TA is easy to contract volitionally but it is affected by volume conduction from the muscles innervated by the tibial nerve. In contrast, the AH is hardly affected by volume conduction from the muscles innervated by the peroneal nerve, but it is difficult to contract only this muscle. Therefore, the recorded muscle should be selected based on the patient’s clinical features.

SAFETY

Magnetic stimulation should be performed according to safety guidelines published by the International Federation of Clinical Neurophysiology.⁸⁰,⁸¹,⁸² Serious adverse effects are extremely rare in single-pulse magnetic stimulation, and it is generally considered safe. However, this technique is contraindicated in patients with metal devices (e.g., cardiac pacemakers, deep brain stimulation electrodes, and cochlear implants), as well as in patients with serious heart disease and in pregnant females. Care should also be taken in patients with a high or unknown risk of seizures, including in those with a history of epilepsy, taking drugs that lower the seizure threshold, sleep deprivation, or alcohol abuse.

The sound associated with the TMS pulse when using high stimulus intensities can exceed the maximum safe level recommended for the hearing system, and so subjects should wear earplugs during the examinations. This is particularly important when using a double-cone coil during stimulation at the foramen magnum, since the coil edges are close to the ears. Additionally, stimulation causes contraction of the masseter muscles due to activation of the cranial nerves, and so care must be taken to ensure that the patient does not bite their tongue. The heart can be influenced when using a MATS coil to stimulate the cervicothoracic region of the back, and so this type of coil should be only used on the lumbosacral region. Emergency treatment equipment, such as monitoring systems and first aid carts, should be available in case of unexpected events. Performing this work where a hospital is nearby is recommended in case emergency events occur.

To perform TMS effectively and safely, the International Federation of Clinical Neurophysiology has developed training guidelines for magnetic stimulation.⁸³ The Japanese Society of Clinical Neurophysiology has also initiated training programs. We recommend that neurologists participate in such training programs and also encourage that they learn directly under the guidance of experts.

CONCLUSION

Single-pulse magnetic stimulation is the basic method used for TMS. Although not mentioned in detail here, there are several stimulation methods other than single-pulse TMS (e.g., paired-pulse TMS and rTMS). Even when performing other types of TMS, the technique of single-pulse TMS is also always required to evaluate the target effects. Single-pulse magnetic stimulation is a useful method for diagnostic neurophysiological examinations of the central and peripheral nervous systems. We propose that all neurologists should learn to perform single-pulse magnetic stimulation since it is a clinically useful technique.

Availability of Data and Material

Data sharing is not applicable to this article since no data sets were generated or analyzed during the described study.