The Origins of Neuromuscular Electrodiagnosis, 1800–1950: A Crucial Period

At the beginning of the 19th century, the manufacture of batteries revived interest in medical electricity. The first version of these batteries was developed by the Italian Alessandro Volta (1745–1827). The improved versions developed by the English physicist John Frederic Daniell (1790–1845) and the German chemist Robert Wilhelm Bunsen (1811–1899) produced a homogeneous electric current. Batteries were incorporated into electricity-generating machines for medical use. This use was known as galvanism, named after the Italian scientist Luigi Galvani (1737–1798). The discovery of electromagnetism and induction currents also contributed to the renewed interest in medical electricity. It stemmed from works by the Danish physicist Hans Christian Ørsted (1777–1851), the French physicist André-Marie Ampère (1775–1836), who formulated the first laws of electrodynamics, and the British physicist Michael Faraday (1791–1867), who theorised the “electromagnetic field”.

The first electromagnetic machines were formed of a magnet and a coil that could be moved relative to each other. They were built in the 1830s by the Frenchman Hippolyte Pixii (1808–1835), then by the Irishman Edward Marmaduke Clarke (1791–1859). In 1851, Heinrich Daniel Ruhmkorff (1803–1877) created the first induction coils, which quickly became a great success. The “Ruhmkorff coils” were included in a large number of electrical devices made by specialist manufacturers and were proliferated throughout Europe.

At that time, most electrotherapists refused to rely on electrophysiological research to advance their practices, which could then only be explained by the “magic aspect” of electricity. The first fundamental works on electricity were based on animal experiments; only a few experiments with electrical stimulation on humans were carried out at the turn of the 19th century. These were often electrical self-stimulations, such as those undertaken by the German scientist and explorer Alexander von Humboldt (1769–1859) and the German physician Johann Christian Reil (1759–1813), who experimented on themselves to investigate the muscular effects produced by electrical nerve stimulation [2, 3]. The Italian Giovanni Aldini (1762–1834) also developed some experiments using human nerve stimulation [4].

In 1822, the French physician François Magendie (1783–1855) used galvanism to distinguish between the functions of the ventral and dorsal roots of the spinal nerves [5]. His collaborator, the physician Jean-Baptiste Sarlandière (1787–1838), developed electropuncture – intra-tissular electrical stimulation with a needle, designed to potentiate acupuncture with electricity [6].

After these initial works, in the early 1840s, the Italian physiologist Carlo Matteuci (1811–1868) used new measuring instruments to record the electrical current generated by muscle tissue at rest [7]. His work influenced the German school of neurophysiology led by Johannes Peter Mueller (1801–1858) and his students Emil Heinrich du Bois Reymond (1818–1896), who described the nerve action potential, and Hermann von Helmholtz (1821–1894), who first measured the speed of nerve impulses in frogs. The French physician Claude Bernard (1813–1878) studied the action of curare during nervous and muscular stimulations in frogs [8].

It was then recognised that interrupted electrical current had a more powerful effect on muscle contractility than permanent current. The physiologist Eduard Pflüger (1829–1910) used this characteristic to establish the law of “muscle twitches” (Zuckunggesetz) in 1859. He observed in animals the strength, duration, and timing of contractions induced by the opening and closing of the electric current [9]. This law was a fundamental discovery of electrophysiology but is of limited clinical interest. To reproduce it in humans, excessively strong electrical currents were required. These new theories, which were likely to provide a physiological basis for the action of the electric current, were coldly received by electrotherapists, who severely criticised them for several decades: “We vigorously refuse to admit that the works of Dubois-Reymond, Erb, and Ziemssen have served to advance electrophysiology. We instead believe that these heavy and indigestible theories, advanced notably by the first of these authors, have had a deplorable effect” [10].

In the mid-19th century, among the electrotherapists, Guillaume Duchenne de Boulogne (1806–1875) was one of the leading physicians interested in electrodiagnosis. Duchenne was a fine clinician with a mastery of photographic and electrological techniques. He was one of the first to approach the question of medical electricity from the triple angle of physiology, diagnosis and therapy. Seeking to develop a more focused form of electrification, he realised the limitations of Sarlandière’s electropuncture. He therefore designed his own electrical equipment and made “rhéophores,” specific electrodes of various sizes and shapes covered with moist pads. Duchenne observed that the technique using induction current, which he called faradisation, was the most suitable for his localised electrification work [11, 12]. To describe the electrical response of nerves and muscles, Duchenne proposed the term neuromuscular contractility to combine the notions of irritability and motricity. At that time, irritability was defined as the direct action of an electric current on the muscle. It was called “Hallerian irritability” in reference to the work of the Swiss scientist Albrecht von Haller (1708–1777). The term motricity, proposed by the physiologist Pierre Flourens (1794–1867), reflected the contraction of a muscle produced by the excitation of a nerve trunk.

Using direct faradisations of muscles, Duchenne clarified the functional anatomy of many muscles and modified the laws laid down by the English physiologist Marshall Hall (1790–1857). Duchenne’s conclusions pointed towards a reduced electro-muscular contractility in spinal cord and peripheral nerve disorders and normal contractility in cerebral disorders and rheumatic or hysterical causes [11, 12] (Fig. 1).

Duchenne’s methodology and conclusions led to a heated debate with the German physician Robert Remak (1815–1865), who advocated the use of galvanic current [13, 14]. The question of the preferred points of muscle stimulation, called “points d’élection” (chosen points) by Duchenne, was also discussed. Duchenne’s study of facial expressions using electrical stimulation of the facial muscles, which was immortalised in photographs, contributed to his renown. Hugo von Ziemssen (1829–1902) also took an interest in these questions and introduced the idea of the immersion point, defined as the point of entry of the motor nerves into the muscle. He studied these points on dying patients and by cadaveric dissection immediately post-mortem. Von Ziemssen proposed didactic diagrams explaining the location of these points, which were to become the “motor points,” allowing the best positioning of electrical stimulation [15].

A return to favour of galvanic current led to the polar method, a new electrical stimulation technique. Jean-Baptiste Auguste Chauveau (1827–1917), a veterinary anatomist, demonstrated the interest of this method in animal physiology, and the German Rudolf Brenner (1821–1884) described its usefulness in humans [16, 17]. The polar method required a large electrode, usually placed on the sternum, and a smaller electrode placed on the nerve, which delivered excitation via the negative or positive pole of the battery. With this method, which gradually became more widespread, the strongest excitation produced by the electricity arises when the galvanic current terminates at the negative pole, and the weaker arises from the positive pole.

A literal notation of the polar method was formalised by Brenner, using the terms anode (An) to define the electrode connected to the positive pole of the battery, and cathode (Ka) for that connected to the negative pole. The closing of the electric current was abbreviated by S (Schliessung), the opening by O (Oeffnung), and the contraction obtained by Z (Zuckung). The muscle contraction formula could thus be expressed by a system of letters. For example, a muscle contraction when the negative pole closes was abbreviated to KaSZ.

In the years following the Franco-Prussian War of 1870–1871, this notation became a subject of debate in France. German ideas were not welcomed at that time and the French physician Joseph Grasset (1849–1918) suggested using abbreviations of French words, for example, F for fermeture (closing), which undoubtedly led to confusion with the German notation [18]. Romain Vigouroux (1831–1911), electrotherapist at the Salpêtrière hospital, disputed Grasset’s suggestions: “There is therefore really no reason not to keep Brenner’s original notation, unless we bring in national sentiment, which has no place in such matters.” [19].

From the 1880s onwards, although electrotherapy often remained the primary motivation for research, a standardisation of electrodiagnostic exploration began to emerge. It laid the foundations for a reasoned diagnostic use of electric current by recommending (1) comparison of the patient’s unaffected side with the affected side and (2) the establishment of standards. The abnormalities observed during electrical stimulation were now described as quantitative, to define an increased or decreased reaction, and qualitative, when they concerned the mode and appearance of reactions.

The German neurologist Wilhelm Erb (1840–1921), in Leipzig, was one of the propagators of these new principles. He also formalised a more complex situation linked to the electrical stimulation of nerves and muscles: Entartungsreaktion (degenerative reaction). This corresponds to a loss of faradic and galvanic excitability of the nerves and faradic excitability of the muscles, while the galvanic excitability of the muscles is preserved, usually with changes in quality and quantity. These findings had already been mentioned in previous decades, in particular by Eduard Baierlacher (1825–1889). Erb gave a precise chronology according to the severity of the damage, opening the way for future electro-clinical correlations [20]. These first potential correlations with the anatomical nerve lesions described by Louis-Antoine Ranvier (1835–1922) and Augustus Waller (1816–1870) were taken up and discussed, particularly in the work of the French physician Aimé Estorc (1855–1888) [21].

Erb was also interested in the question of the human body’s resistance to conductivity and advocated a study of this resistance for the different regions of the body. The Swiss neurologist Paul Dubois (1848–1918) also took an interest in this question [22]. The second international congress on electrotherapy, which he organised in Bern in 1902, devoted a great deal of attention to the subject.

In 1909, the physiologist Louis Lapicque (1866–1952) at the Sorbonne University in Paris defined rhéobase (rheobase), the electrical threshold required to trigger neuronal excitation, and chronaxie (chronaxis), the time required to reach rheobase twice [23]. These parameters were quickly adopted for measuring neuromuscular function in laboratory animals. Georges Bourguignon (1876–1963) had great difficulty applying the concept to humans because of difficulties in measuring.

Another recurring problem was that of quantifying the intensity of the electric current used. Simply counting the number of battery cells used had become insufficient. For faradic stimulation, the distance of insertion of the primary coil into the secondary coil, measured in millimetres, was recorded. For galvanic stimulation, the deviation of the galvanometer needle, often expressed in degrees, was used. The definition of the Ampere unit finally standardised the graduation of galvanometers in milliamperes.

The use of medical electricity during the First World War (1914–1918) was marred by misuse and medical-military abuses for the “treatment” of wartime psychoneurosis [24]. However, the war also saw the widespread use of electrodiagnosis due to the frequency of nerve injuries. Adolphe Zimmern (1871–1935) and Pierre Pérol (1880–1961) proposed a standardised diagnostic strategy [25]. The bipolar method and the development of galvanometers improved the quantification of electrodiagnosis results.

Physicist and physician André Strohl (1887–1977) developed the first electrical recordings of tendon reflexes, following on from the work of the German physician Paul Hoffmann (1884–1962), who in 1910 described the recording of the reflex that now bears his name [26]. Strohl recorded the reflexes of the military patients studied by Georges Guillain (1876–1961) and Jean-Alexandre Barré (1880–1967) for the description of the syndrome that would become known as Guillain-Barré syndrome [27].

The study of the electrical activity of muscle, now called myography, underwent significant progress in the first decades of the 20th century thanks to physiological advances and major technical innovations (Fig. 2). Physiological work on animals gradually led to the concept of the motor unit. In 1909, in Cambridge, the British scientist Keith Lucas (1879–1916) used the model of the neuromuscular system of the frog to observe that the muscle fibres innervated by a nerve were activated according to the “all-or-nothing” law [28]. The concept of the motor unit, in which a motor neuron and its axon innervate a group of muscle fibres in a specific muscle, was finally established in 1925 by Edward George Tandy Liddell (1895–1981) and Charles Scott Sherrington (1857–1952), chair of physiology at Oxford University [29, 30]. This concept stated that, during a muscular contraction, the motor units are activated independently of each other. Recording the electrical activity of the muscle must therefore focus on the motor unit potentials.

The major technical discoveries, such as more sensitive recording equipment, helped achieve this objective. Edgar Douglas Adrian (1889–1977), who took over Keith Lucas’s laboratory in Cambridge, collaborated with the American physiologist Detlev W. Bronk (1807–1975) to design a concentric needle electrode made from copper wires in a hypodermic needle. This new intramuscular sensor led to the selective recording of the activity of a few muscle fibres, helped by the use of a loudspeaker to amplify the sound of muscle electrical activity [31]. In 1897, the cathode-ray tube was developed by Karl Ferdinand Braun (1850–1918) [32]. It overcame the inertia of the various galvanometer recording systems and gave rise to the cathode-ray oscilloscope. In 1922, Joseph Erlanger (1874–1965) and his student Herbert Gasser (1888–1963) from Washington University in Saint Louis successfully used Braun’s cathode-ray oscilloscope to graph the action potential of a frog’s sciatic nerve [33]. They established a first clear relationship between the size of nerve fibres and the speed of nerve conduction. Bryan HC Matthews (1906–1986), Adrian’s colleague [31] at Cambridge University improved recording capabilities by using an amplifier and an oscillograph. In 1934, he developed a differential amplifier that permitted more efficient electrophysiological measures [34].

These major physiological and technological advances made it easier to record motor unit potentials. The first myographic descriptions of pathological situations were proposed in the 1930s. In 1935, Donald Benjamin Lindsley (1907–2003), an American physiologist, studied the myographic activity of patients suffering from myasthenia and noted a reduction in the amplitude of motor unit potentials during muscle activity [35]. In 1938, Derek E. Denny-Brown (1901–1981) and Joseph Buford Pennybacker (1907–1983) published a study of motor unit action potentials in patients suffering from neurological disorders. This publication is one of the founding articles of electro-clinical semiology. In patients with amyotrophic lateral sclerosis, the two authors were able to separate the action potentials of muscle fibres innervated by a damaged nerve, corresponding to fibrillation in its current definition, from the action potentials of abnormal motor neurons, now known as fasciculations [36]. In 1941, Denny-Brown, who continued this work, described the electrical characteristics of myotonia [37].

The rise of Nazism and the Second World War turned medicine towards military applications and changed the way hospitals and laboratories operated. German doctor Fritz Buchthal (1907–2003), working at the Kaiser-Wilhelm-Institut für Biologie, was forced to move to Denmark to escape the Nazis. He continued his study of human myography, working on the differentiation of muscular atrophy using a needle electrode [38].

The number of nerve injuries caused by the various firearms used during the conflict made it essential to have a reliable tool capable of locating the site of nerve damage and assessing its severity. The US Office of Scientific Research and Development, created in 1941 after the USA entered the war, commissioned several researchers to work on this issue. One of them was the neurologist James G. Golseth (1912–2003) from Percy Jones General Hospital in Battle Creek (Minnesota) collaborated with the engineer James A. Fizzell (1912–1995) to develop a constant-current pulse stimulator for measuring the electrical activity of the nerve (Fig. 3).

Golseth and Fizzell got in touch with Herbert H. Jasper (1906–1999) of McGill University in Montreal, who was also working on the electrical study of injured nerves in Canadian soldiers. An active collaboration was established. Golseth and Fizzell carried out nerve conduction studies and Jasper performed myography. The benefits of pooling and interpreting these two techniques together quickly became obvious. The three researchers aimed to develop electromyographic and electrodiagnostic equipment that could be used in everyday clinical practice [39]. Jasper and the biophysicist André Cipriani (1908–1956) developed a portable electromyograph. In 1953, Golseth became the first president of the American Association of Electromyography and Electrodiagnosis.

In Oxford, England, the orthopaedic surgeon Herbert Seddon (1903–1977) set up a team to study and repair nerve damage caused by war injuries. In 1942, he proposed a now classic classification into three types of nerve injury: neurapraxia, axonotmesis, and neurotmesis [40]. Seddon called on the services of Graham Weddell (1908–1990). Weddel served in the Royal Army Medical Corps in Oxford. He had an electromyography (EMG) machine of his own making in his anatomy laboratory. With Bertram Feinstein (1914–1978) and Richard Pattle (1918–1980), he described the evolution of muscle denervation. Their work demonstrated the clinical value and chronology of appearance of fibrillation potentials in the event of nerve damage and defined reinnervation potentials [41].

This seminal work, which emerged during the Second World War, was the fruit of collaboration between doctors, physiologists, and physicists. Some of these pioneers left the neuromuscular field after the war. Graham Weddell became a leprosy specialist and Richard Pattle became a pulmonary surfactant specialist. Their intensive collaboration led to the design, in different laboratories, of the first machines combining electrostimulation and myography. This late combination of needle EMG and measurements of nerve conduction velocities gave rise to terminological hesitations between the term EMG, which refers only to the study of needle myography, and the term ENMG, which refers to the study of nerve conduction and myography.

In the 1950s, the first widespread marketing of electromyographic devices combining stimulation and myography technologies appeared. From the end of the Second World War onwards, a generation of neurologists, such as Fritz Buchthal in Denmark, Eric Kugelberg (1913–1983) in Sweden, François Isch (1918–2004) in France, Roger W. Gilliatt (1922–1991) in the UK, and Edward H. Lambert (1915–2003) in the USA, invested more intensively in this new diagnostic technique and established the standards and premises of electro-anatomical correlations on a larger scale. The subsequent progress, which led to the development of EMG as we know it today, came after the introduction of microelectronics and computing.

We thank Jennifer Dobson for the proofreading of the manuscript and the translation into English of the French quotations.

An ethics statement was not required for this study type since no human or animal subjects or materials were used.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

L.T.: data collection, literature search, and writing of the manuscript. Y.P.: writing and editing the manuscript.

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.