Summary

Neurophysiological procedures are important investigations to diagnose or guide the management of various neurological and neuromuscular diseases. Tests commonly available include electroencephalography and electromyography; while video electroencephalography and evoked potential studies usually have more specialized applications. The role of clinical neurophysiology has changed tremendously in the last decade because of advancement in medical technology and reforms of the healthcare system. In this article, we shall introduce some basic concepts of neuro-physiological procedures and provide an update on their applications in clinical practice.

摘要

神經電生理學主要應用於診斷各種中樞及周邊神經系統和肌肉的疾病及指導其治療。它藉著測量神經及肌肉運作時產生的生理電流變化，而對其功能作出客觀的評估。常用的神經電生理學檢查包括腦電圖和肌電圖，而錄影描記遙測腦電圖和神經誘發電位則用於特定範圍。隨著醫療科技及制度的發展，神經電生理學在臨床醫學上已重新定位。本文將探討神經電生理學的基本概念及其在現今臨床醫學上的應用。

---

Introduction

Electroencephalography (EEG) and electromy-ography (EMG) are the commonly performed neurophysiological procedures. Other tests that are available upon request include video EEG and the various evoked potential (EP) studies. In our laboratory, over 3,000 procedures are carried out every year.

Activities of excitable tissues, including nerves and muscles, are associated with membrane potential changes that generate minute electrical signals. Signals from respective regions of the neuronal system can be recorded with combinations of filters and amplifiers to select the suitable frequency and voltage ranges. EEG and EP assess the central nervous system while EMG is for the peripheral neuromuscular system. By recording the electric activities of these excitable tissues, their functional status can be assessed objectively. Data generated is usually quantitative and reproducible. The majority of procedures in routine clinical use are non-invasive.

Clinical neurophysiologists are physicians who have undergone training in the diagnosis and treatment of neurological and neuromuscular diseases and the application of electrophysiological techniques to study these disorders.¹ Physicians running a neurodiagnostic laboratory are familiar with the instrumentations, biomedical electronics, quantification and statistical analysis of electrophysiological data, as well as basic neural and neuromuscular anatomy, physiology, and pathology. They are also accredited in neurology or related specialties. Knowledge in the clinical aspects of neurological and neuromuscular diseases is pertinent to the proper performance of electrophysiological evaluations.²

Recent advances in medical technology have expanded the scope of clinical neurophysiology. Reforms of the healthcare delivery system have also changed its role in medical practice.

Electroencephalography

EEG is indicated in ALL patients with possible seizure disorders or known to have had epilepsy. It is important to classify seizure disorders³ and hence determine the mode of treatment. It is also an important part of investigating causes of possible seizure disorders. EEG study records and analyses electrical signals generated by the cerebral cortex. These signals are small signal). Moreover, they are surrounded by a variety of large electrical potentials from the environment (noise). Although good equipment and meticulous technique can improve signal-to-noise ratio, a good or clinically useful EEG study also needs an informed interpretation of data. These factors are equally important for proper use of this neurodiagnostic tool.

Physiological principles of the EEG

EEG signals are generated by the cerebral cortex. Spontaneous EEG activity reflects the current flows in the extracellular space. This is the summation of excitation and inhibition synaptic potentials occurring on thousands or even millions, of cortical neurons. Individual action potentials do not contribute directly to EEG activities.

Conventional EEG is a continuous graph over time of the spatial distribution of changing voltage field at the scalp surface. EEG also depends on afferent inputs from subcortical structures, e.g., thalamus and brainstem reticular formation. Since EEG is not the same as electrocorticogram (see below), not all potentials produced at cortical surface are detectable on scalp EEG. Location, voltage, and extent of epileptiform discharge affect the detection of abnormalities by scalp EEG study.

EEG recording technique

An interictal EEG study lasts for about 45-60 minutes. The patient is connected by a series of small gold, sliver or silver-silver chloride discs symmetrically over the scalp on both sides, in accordance with the international 10-20 system.⁴ In practice, 16-20 channels of EEG activity are recorded simultaneously. A period of spontaneous brain wave activity is recorded, followed by activation procedures, namely hyperventilation and photic stimulation, to enhance detection of abnormalities. EEG may be performed under sleep or after sleep deprivation. Additional electrodes (e.g., sphenoidal electrode) may be used in specific situations. Figure 1 illustrates a commonly used electrode placement for routine EEG recording.

Normal EEG activities (Figure 1)

Spontaneous fluctuations of voltage potential at the cortical surface are in the 100-1000 mV range but are only 10-100V at the scalp. Different parts of the cerebral cortex generate relatively distinct potential fluctuations, which also differ in the waking and sleep states.

In normal adults, the waking EEG pattern consists of sinusoidal oscillations at 8-12 Hz which is most prominent over the occipital area (alpha rhythm). Such alpha rhythm is attenuated (or blocked) by eye opening, mental activities, and drowsiness. Activity faster than 12 Hz (beta activity) is normally present over the frontal area. It may be more prominent in patients receiving barbiturates or benzodiazepines. Activity slower than 8 Hz is subdivided into delta activity (1-3 Hz) and theta activity (4-7 Hz). A small amount of theta activity over the temporal regions can be considered as normal. The percentage of intermixed theta activities increases after the age of 60. Delta activity is not normally found in awake adults but may appear when they fall into sleep. The amount of slow activity (theta and delta) correlates closely with the depth of sleep. Slow activity is abundant in the EEG of newborns and young children, but disappears progressively with maturation.

Digital EEG

Computer-based EEG machines are now readily available. The term "digital" refers to the fact that the acquired EEG information is stored, processed and reproduced by a computer. The analog EEG signals are transformed into a series of numbers that specify the original signals sequentially at short time intervals - a process termed analog-to-digital conversion. Digital EEG has several advantages over analog recording: 1) Digital EEG can be transmitted to networked review stations for either off-line or on-line analysis; 2) The recorded signals can be reformatted retrospectively using different time scales, filters or montages for optimal display of abnormalities; 3) The recording and display procedures are essentially paperless; 4) Innovative display methods can be used to present EEG results to non-electroencephalographers; and 5) Storage of data is much more effective than the conventional paper system (a computer file on CD or hard disk versus a pile of paper for each recording).

Clinical uses of EEG (Table 1)

EEG assesses physiological alterations of brain activities. Although many changes are non-specific, some are highly suggestive of specific entities (e.g., epilepsy, herpes encephalitis, metabolic encephalopathies). EEG is also useful in following up the course of patients with altered consciousness and provides prognostic information, including determination of brain death.

EEG is a tool to answer a specific clinical question. Therefore, sufficient clinical information is crucial for the electroencephalographer for an informed EEG interpretation.

EEG is the most useful diagnostic test when epilepsy is being suspected. Because the onset of seizure is unpredictable and their occurrence is relatively infrequent in most patients, EEG is usually obtained during the interictal phase (in between seizures). Electrical abnormalities may be detected in patients with epilepsy even between attacks. However, interictal findings must be interpreted with caution. Most epileptiform discharges correlate poorly with the frequency or likelihood of seizure recurrence. The most informative or diagnostic scenario is to obtain an EEG recording during a typical seizure attack. Such recording is now possible by performing prolonged video-EEG monitoring. EEG abnormalities could be focal (e.g., temporal lobe epilepsy) or generalized (e.g., absence seizure). The type of epileptiform activity is crucial for seizure classification and, sometimes, for identifying a specific epileptic syndrome. Nevertheless, a normal EEG does not exclude epilepsy, and an abnormal EEG does not necessarily indicate epilepsy. The sensitivity of detecting epileptiform activity by a single awake interictal EEG recording is about 50% in adults with epilepsy, while sleep study increases the sensitivity to 80-85%. Two awake recordings will demonstrate epileptiform activity in 85% of individuals and this rises to 92% with four recordings.⁵ On the other hand, epileptiform activities could be found in 2% of non-epileptic subjects.

EEG is an important diagnostic and monitoring tool for patients with altered state of consciousness. It complements clinical examination and neuroimaging studies. Although abnormalities are typically non-specific with regard to aetiology, there is a good correlation with the clinical state. EEG can help to answer the following questions:

1. Are psychogenic factors playing a major role?
2. Is the on-going process diffuse, focal or multifocal?
3. Is epileptic activity depressing consciousness (e.g., non-convulsive status epilepticus)?
4. Is there evidence of improvement despite relatively little clinical change?
5. What is the prognosis?

Metabolic or toxic encephalopathies, hypoxia, infectious diseases and suspected brain death are the most frequent clinical indications for EEG in patients with altered consciousness. For management of patient with status epilepticus, either convulsive or non-convulsive, interval EEG examination or continuous EEG monitoring are invaluable.

With all infectious diseases affecting the brain, EEG is most useful in the initial assessment with possible herpes simplex (HSV) encephalitis. The characteristic EEG changes in HSV encephalitis help select patients for early treatment and biopsy. Non-HSV encephalitis cause diffuse polymorphic slow wave activity, which is non-specific and can also occur in other encephalopathies. In contrast, a normal EEG makes the diagnosis of encephalitis unlikely.

EEG may be useful for assessing patients with dementia. However, due to problems encountered in distinguishing the effects on cerebral electrical activity of normal aging from those caused by disease processes and absence of generally accepted quantifiable methods of analysis and statistically valid comparison measures, its use is rather limited. In practice, EEG can supplement the evaluation of suspected dementia by revealing abnormal cerebral function where there is the possibility of a psychogenic disorder, and by delineating whether the process is focal or diffuse. An example is identifying the typical periodic sharp wave complexes in Creutzfeldt-Jakob disease presenting with dementia.

EEG was an important diagnostic tool for detecting focal cerebral lesions. However, with the advent of neuroimaging technology, this role has declined.

Video EEG examination

Video EEG (VEEG) examination can be considered in selected patients with seizure disorders. VEEG is a time-locked, synchronized EEG and video recording. It provides us an opportunity to examine seizure semiology. As localization and preferred route of electrical spread determine the clinical features, critical analysis of the seizure pattern may localize the epileptogenic focus, for instance, oro-alimentary automatism of temporal lobe epilepsy. It is also useful in differentiating pseudoseizure from epileptic seizure when the diagnosis of epileptic seizure cannot be ascertained from the patient's history.

The localization value of ictal scalp EEG had been studied;⁶ it could discriminate out 70, 52, 23 or 10% of occipital, lateral temporal, frontal or parietal foci, respectively. If surface recording fails to demonstrate the seizure origin in a potential surgical candidate, depth or subdural intracranial electrodes (electrocorticography) may be considered for defining epilepsy onset and its route of propagation. Although this procedure carries some morbidity, mortality is seldom reported. It can be safely performed by experienced neurosurgeons and guides the extent of surgical resection for improving seizure control while preserving the eloquent cortex.⁷

Electromyography and nerve conduction studies

The term "EMG" is often used to encompass the entire spectrum of peripheral electrodiagnostic studies. The two main components are nerve conduction studies for large fibre functions and needle electromyography examination (NEE) for bioelectric activities of skeletal muscles. Their indications are summarized in Table 2. Other peripheral techniques commonly used include thermal threshold for testing small fibres, sympathetic skin response for autonomic function, and repetitive nerve stimulation and single-fibre EMG for evaluating neuromuscular junction. Axonal viability during the acute phase of neuropathies, such as Bell's palsy, can be assessed by nerve excitability tests.

Nerve conduction studies

Depolarization of a peripheral nerve can be induced by a small electric stimulation. The current used is less than 100 mA for 0.1-1 msec, which only causes minor discomfort. The propagating membrane action potential after nerve depolarization can be recorded either directly over sensory or mixed nerves or using a muscle for motor nerves. Stimulation and recording points are located by surface landmarks. Conduction velocity in a nerve segment is estimated by the latency difference of action potentials from two stimulation points. Amplitude of an action potential reflects the number of axons in the nerve being tested. Disorders affecting the nerve myelin, such as the inflammatory and entrapment neuropathies, produce conduction slowing, while problems causing loss of axons result in attenuation of action potential amplitudes. Nerve conduction parameters obtained are compared against a reference range, which can either be published normative data¹⁰ or developed by the laboratory from its own controls.¹¹ A whole range of nerves can be studied, though a lot are not routinely tested (Table 3). Proximal parts of nerves not accessible with surface stimulation can be tested by the late responses, including F-waves through the motor axon and H-reflexes through the spinal reflex arc, but they are not sensitive for focal lesions.

Needle electromyographic examination

The peripheral neuromuscular system is made up of motor units. A motor unit begins at the anterior horn cell in the ventral spinal grey matter. The anterior horn cells are grouped in nuclei-like structures for their respective innervated muscles. The motor neuron then courses along the spinal root, plexus and peripheral nerve, and ends in the muscle fibres through the neuromuscular junction. Muscle fibres innervated by different motor units are mixed within the same muscle. The mean number of muscle fibres per motor unit ranges from 100-400 in small hand muscles and 600-2000 in large limb muscles. The motor unit territory is roughly 5-7 mm in the upper and 7-10 mm in the lower limbs. A motor unit is made up of subunits, which contains 10-30 muscle fibres. When a motor subunit is activated, all its muscle fibres will fire simultaneously. NEE is performed by inserting a needle electrode into the muscle being tested. The patient is asked to contract the muscle with the slightest effort. The aim is to activate individual motor units. A concentric needle electrode has an external diameter of 0.3-0.65 mm, which is much smaller than a standard venepuncture needle. The recording area is between 0.5-1 mm². Signals within this area will summate to form motor unit action potentials. Pathologies in nerves and muscles can be reflected by changes in their configurations (Figure 2) and recruitment patterns. Software for the quantitative analysis of these EMG data are available. Other important EMG abnormalities include fibrillations, fasciculations, and continuous motor activities such as myotonia, myokymia and complex repetitive discharge.

The usefulness of NEE is very dependent on the electromyographer's skill. Interpretation of NEE findings takes place alongside with the test. There is no set protocol for choice of muscles to be sampled. Each NEE is individually designed, basing on the clinical circumstance and information obtained during the study. Therefore, NEE cannot be delegated to a technician. NEE carries the risk of injury to nerves and blood vessels and must be performed by a physician with adequate knowledge in anatomy. Sampling of deeper muscles or within a fascia compartment is contraindicated in patients with bleeding tendencies (e.g., taking anticoagulants). Precautions should also be taken to avoid transmission of infections through needle-prick injuries. Finally, NEE will not be successful unless the patient is able to cooperate and tolerate the procedure.

Useful references for electromyographers include anatomical guides and atlases.^12-14 A demonstration of some of these techniques is also available on video.¹⁵

The electrodiagnostic consultation

Electrodiagnostic (EDx) consultation is the comprehensive assessment and diagnosis of a clinical problem that suggests a neuromuscular disorder.¹⁶ The essential components include defining the clinical problem, selection of relevant parameters to be collected, interpretation of data, and their integration into a diagnosis. It should be completed with a recommendation on management or some objective prognostic information. Nevertheless, EDx studies are not a substitution for careful history taking and physical examination. A consultation takes 30 minutes to over two hours, depending on its complexity. Useful guidelines, including indications for referral, are available on the American Association of Electrodiagnostic Medicine (AAEM) website, www.aaem.net/aaem/PracticeIssues/PIIndex.cfm.

There are two fundamental but contradictory approaches to EDx consultation. Some consider electrophysiological studies as an extension of the neurological examination and a full clinical evaluation by the electromyographer should be performed beforehand. Others advocate that these tests should be conducted by an independent physician and stand as a second opinion. Too detailed a clinical assessment might bias its neutrality. Different laboratories adopt different approaches but, as the minimal requirement, a focused neuromusculoskeletal history and physical examination should be performed to define the possible differentials before proceeding to electrophysiological studies. The subsequent process is a hypothesis-testing exercise of the initial diagnostic assumptions. These principles of clinical EDx reasoning were reviewed by Campbell.¹⁷

"Shotgunning" and "focused searching" are the two conceptually different strategies in data collection. The former is more "mechanical" in which a standard set of parameters is routinely gathered regardless of the clinical problem. This approach is practiced by laboratories run by less experienced operators or technicians. Academic institutes also apply standardized protocols to collect research data. Focused searching is a more interactive approach in which each examination is "tailored" to the patient's clinical circumstance. Data relevant to the presenting problem is collected, continuously monitored and interpreted during the examination. What needs to be covered depends on the findings as they unfold. Non-protocol tests may be included. This approach relies heavily on the electromyographer's experience and logistic skills. For more complex clinical problems in which meaningful diagnostic conclusions cannot be drawn from routine data collection, the focused searching approach is more appropriate.

Sometimes, a patient is referred for a specified test, such as NEE of one particular nerve root or muscle. This is usually initiated by a physician familiar with the indications of electrophysiology and looking for a specific answer. It is often unnecessary to go beyond what is requested. Patients may also be referred for "ruling out" a particular problem, like carpal tunnel syndrome. If no electrical evidence is found to support the diagnosis, one can either stop the study, having ruled out the problem as requested, or proceed to further tests for other conditions that might explain the symptoms, such as cervical radiculopathy. The latter "rule-in" approach is more helpful for patient management.

A misleading EDx report may subject a patient to unnecessary invasive interventions. Overcall of borderline electrophysiological abnormalities should, therefore, be avoided. Conversely, missing potentially treatable conditions is equally harmful. A competent electromyographer should be aware of the consequences of improper data interpretation and reporting as well as the limitations of EDx procedures. It is sometimes appropriate to define normality by observing EDx criteria, such as those published in the AAEM Practice Parameters. However, these guidelines are not always straightforward to use and may have poor sensitivity. In many conditions, they are research criteria rather than adapted for routine clinical application. Another pitfall is incomplete information gathering. Omission of essential tests can be due to poor knowledge or mere carelessness. Similarly, drawing premature conclusions without careful consideration should be avoided. If there is an indication, the patient can be assessed again for reproducibility of the initial findings and to look for progression of deficits. Other limitations include lesions that are not technically accessible, sampling error, wrong timing of test, sensitivity problem, etc., such that a normal study is not equivalent to absence of neuromuscular disease. Therefore, when reporting on the EDx findings, it is often appropriate to accommodate for these limitations by addressing a degree of uncertainty. Moreover, electrophysiological impairments are seldom pathognomic of specific conditions, although a pattern of abnormality may be recognized to give us diagnostic clues.

Evoked potentials

EP is a functional technique for lesion localization. Various components of the neuraxis can be studied; the somatosensory, visual, auditory and motor pathways are most frequently tested.

When the dorsal column is activated by stimulating a nerve in the upper or lower limb, instead of a monophasic response, several consistent peaks of signal can be recorded over the afferent pathway. These are called the somatosensory evoked potentials (SEP). Brainstem auditory evoked potentials (BAEP) or visual evoked potentials (VEP) are elicited by auditory or visual stimuli, respectively.

Evoked potentials are generated by the synchronized activities of neuronal populations, known as neural generators. They can either be anatomical structures like synapses or physical changes in the conduction pathway such as alterations in volume or direction of electric propagation. These neural generators correlate with anatomical landmarks. Prolongation of inter-peak interval of two EP components or delayed latency of a subsequent response indicate a lesion causing conduction slowing in the pathway between their respective neural generators (Figure 3). Abnormal EP is not diagnostic of any specific disease entity. Rather, it reflects the impaired neuronal function from the underlying pathology. Before the imaging era, SEP was used routinely for localizing spinal cord lesions and mechanical entrapments and BAEP for brainstem lesions. However, these applications are largely replaced by magnetic resonance studies. Evoked potentials are now mainly for assessing the physiological functions of conducting pathways, such as in intra-operative monitoring, and determining the extent of multifocal neurological conditions, such as multiple sclerosis or spinocerebellar degenerations.

Motor evoked response is usually obtained by transcranial magnetic stimulation. The motor cortex is activated by a magnetic field-induced current. A volley of depolarization then travels down the spinal cord and motor axons to generate a muscle response. It assesses the integrity of pyramidal and motor conduction pathways.

How to read and use an EDx report

A detailed EEG report consists of four parts: introduction, description, impression and clinical correlation.¹⁸

1. Introduction describes the reason for performing EEG with an appropriate diagnosis, and the electrodes used in the study.
2. The factual EEG findings are outlined, which include any significant background abnormalities, the location of epileptiform discharges, and the most frequent sites of such discharges.3. Impression describes whether the EEG is normal or abnormal, with the abnormal findings listed.4. In clinical correlation, the appropriate clinical diagnoses that are supported by EEG findings are given.

We suggest Family Physicians to focus on the impression and clinical correlation. The structure of an EMG report is similar:¹⁶

1.  
   1. The first part outlines the reason for referral or indication of study, and some key points in history and physical examination.
   2. Secondly, the electrophysiological findings are described. Some laboratories also include a numerical table of the data obtained with their reference values.
   3. The last part gives a diagnostic impression. The degree and significance of electrophysiological abnormalities are commented on. The electromyographer will determine whether the EDx findings can explain the clinical picture. Limitations of the study will be addressed when appropriate. If a clear diagnostic conclusion can be formulated, the electromyographer usually comments on the prognosis or recommends on a treatment option. Conversely, the electromyo-grapher will deduce the differential diagnoses basing on anatomical and physiological abnormalities and suggests further investigations or actions required.

A properly constructed EEG or EMG report will not only aid the clinical diagnosis but also provide useful information to help patient management. Nevertheless, how much an EDx opinion weighs when making treatment decisions should be determined by the managing clinician.

Conclusion

Electrophysiological studies are powerful diagnostic tools for many neurological and neuromuscular disorders. Physicians requesting EEG or EMG should be aware that they are not ordering just a "routine" test run by an automated machine. Integration of relevant clinical information is mandatory for planning the study and interpreting the obtained data. Moreover, limitations of these tests should be recognized; not every aspect of the neurological or neuromuscular system can be covered and there are also diagnostic pitfalls. Therefore, in order to make the best use of these tests, good communication between the referring physician and clinical neurophysiologist is essential.

Key messages

1. Electrophysiological studies (EEG, EP, EMG) assess neuronal functions objectively and quantitatively. Most procedures in routine clinical use are non-invasive.
2. In the hands of trained clinical neurophysiologists, they are powerful diagnostic tools for many neurological and neuromuscular disorders.
3. Abnormal electrodiagnostic findings indicate functional disruption rather than directly tell us the specific aetiological nature of an underlying pathology.
4. Application of electrodiagnostic procedures should be individualised to a patient's presenting problem. The findings must not be interpreted in isolation. Integration with a relevant clinical impression is essential.

---

---

References