Hp2

Geddes, L. A. “The Electrocardiograph.” The Biomedical Engineering Handbook: Second Edition. Ed. Joseph D. Bronzino Boca Raton: CRC Press LLC, 2000

Historical Perspectives 2

The Electrocardiograph

Leslie A. Geddes Purdue University

The First Electrocardiogram Capillary Electrometer Record Rheotome Record Mammalian Electrocardiograms Corrected Capillary Electrometer Records Clinical Electrocardiography The String Galvanometer • Vacuum Tube Electrocardiograph • HotStylus Record

Recording and display of bioelectric events occupied a long time and required the development and adaptation of a variety of primitive instruments, not all of which were electronic. The first bioelectric recorder was the rheoscopic frog, consisting of a sciatic nerve and its innervated gastrocnemius muscle. So sensitive was the nerve that it could be stimulated by the beating heart or a contracting muscle; both events contracted the gastrocnemius muscle. However, such a response provided no information on the time course of these bioelectric events. Sensitive and rapidly responding indicators were essential for this purpose. As Etienne Jules Marey [1885], champion of the graphic method, stated: In effect, in the field of rigorous experimentation all the sciences give a hand. Whatever is the object of these studies, that which measures a force or movement, an electrical state or a temperature, whether he be a physician, chemist or physiologist, he has recourse to the same method and employs the same instruments. Development of the galvanometer and the electric telegraph provided design concepts and instruments that could be adapted to the measurement of bioelectric events. For example, Thomson’s reflecting telegraphic galvanometer was used by Caton [1875] to display the first electroencephalogram. Ader’s telegraphic string galvanometer [1897] was modified by Einthoven [1903] to create the instrument that introduced clinical electrocardiography. Gasser and Erlanger [1922] adapted the Braun cathode-ray tube to enable recording of short-duration nerve action potentials. Garceau [1935] used the Western Union telegraphic recorder, called the Undulator to create the first direct-inking electroencephalograph. However, in the early days of bioelectricity, ingenious electrophysiologists appropriated many devices from physics and engineering to establish the existence of bioelectric phenomena.

The First Electrocardiogram The electric activity accompanying the heartbeat was discovered with the rheoscopic frog by Kolliker and Mueller [1856]. When these investigations laid the nerve over the beating ventricle of a frog heart, the muscle twitched once and sometimes twice. Stimulation of the nerve obviously occurred with depolarization and repolarization of the ventricles. Because at that time there were no rapidly responding galvanometers, Donders [1872] recorded the twitches of the rheoscope to provide a graphic demonstration of the existence of an electrocardiographic signal.

© 2000 by CRC Press LLC

FIGURE HP2.1 The capillary electrometer (a) used by Marey and Lippmann in 1876 and a tortoise ventricular electrogram (b) made with it. This is the first cardiac electrogram from a spontaneously beating heart.

Capillary Electrometer Record The capillary electrometer was created especially for recording the electrocardiogram. The principle underlying its operation was being investigated by Lippmann, a colleague of Marey in France. The phenomenon of electrocapillarity is the change in contour of a drop of mercury in dilute sulfuric acid when a current is passed through the mercury—sulfuric acid interface. This phenomenon was put to practical use by Marey [1876], who placed the interface in a capillary tube, transilluminated it, and recorded the contour change on a moving (falling) photographic plate. The two wires from the electrometer were connected to electrodes placed against the exposed tortoise ventricle. Figure HP2.1a illustrates the capillary electrometer, and Fig. HP2.1b is a reproduction of the tortoise ventricular electrogram showing what we now call the R and T waves.

Rheotome Record Probably unaware that Marey had recorded the cardiac electrogram with the capillary electrometer, Burdon-Sanderson [1879] in England used a slow-speed, d’Arsonval-type galvanometer, the rheotome [see Hoff and Geddes, 1957], and induction-coil stimulator [see Geddes et al., 1989] to reconstruct the ventricular electrogram of the frog heart; Fig. HP2.2 illustrates his reconstruction, showing the R and T waves; note their similarity with those obtained by Marey in Fig. HP2.1b.

© 2000 by CRC Press LLC

FIGURE HP2.2 Burdon-Sanderson’s plot of the frog cardiac electrogram. Thirty-five points were determined to make the reconstruction. [Burdon-Sanderson and Page, 1879.]

Mammalian Electrocardiograms When news of the capillary electrometer reached the United Kingdom, many investigators fabricated their own instruments. One of these was Waller, who used it to record the electrocardiogram of a patient whom he called Jimmy. In 1910, Waller revealed the identity of Jimmy, his pet bulldog, shown in Fig. HP2.3 having his ECG recorded with a forepaw and hindpaw in glass vessels containing saline and metal electrodes.

FIGURE HP2.3 Waller’s patient Jimmy having his ECG recorded with the capillary electrometer. (From Waller AD, Hitchcock Lectures, University of London, 1910.)

© 2000 by CRC Press LLC

FIGURE HP2.4 First human apex cardiogram and capillary electrometer ECG (a) and the dipole map for the heart (b), (a from Waller [1887]; b from Waller [1889].)

Waller [1887] obtained the first ECGs from human subjects; Fig. HP2.4a is one of his records which displays the apex cardiogram and the capillary electrometer record, showing the R and T waves. At that time there were no standard sites for electrode placement. Using the extremities, Waller experimented with different sites, discovering that there were favorable and unfavorable sites, i.e., sites where the amplitude was large or small; Table HP2.1 summarizes his findings. From recordings made with these electrodes, Waller [1887] proposed that the heart could be represented as a dipole, as shown in Fig. HP2.4b.

Corrected Capillary Electrometer Records By the 1890s, it was known that the response of the capillary electrometer was slow, and methods were developed to correct recordings to obtain a true voltage-time record. Burch [1892] in the United Kingdom developed a geometric method that used the tangent at each point along the recording. Figure HP2.5 is an illustration showing a capillary electrometer record (dark hump) and the true voltage-time record (biphasic waveform).

© 2000 by CRC Press LLC

TABLE HP2.1

Waller’s Electrode Locations

The unfavorable combinations were: Left hand and left foot Left hand and right foot Right foot and left foot Mouth and right hand The favorable combinations were: Front of chest and back of chest Left hand and right hand Right hand and right foot Right hand and left foot Mouth and left hand Mouth and right foot Mouth and left foot

One who was very dissatisfied with capillary-electrometer records was Einthoven in the Netherlands. He obtained a capillary electrometer record from a subject (Fig. HP2.6a) and applied the correct method to create the ECG shown in Fig. HP2.6b, revealing the intimate details of what he called the Q and S waves, not visible in the capillary electrometer record, in which he used A, B, and C to designate what he later called the P, R, and T waves.

Clinical Electrocardiography The String Galvanometer Einthoven [1903] set himself the task of creating a high-fidelity recorder by improving Ader’s string telegraphic galvanometer. Einthoven used a silvered quarter filament as the conductor and increased the field strength surrounding the filament by using a strong electromagnet. He added a lens to focus the image of the filament onto a moving photographic surface. Thus the thick baseline for the string galvanometer recording was the image of the “string” (quartz filament). Electrocardiac current caused the silvered filament to be deflected, and its excursions, when magnified optically and recorded photographically, constituted the electrocardiogram. Figure HP2.7a illustrates Einthoven’s string galvanometer, and Fig. HP2.7b shows a patient in hospital clothing having his ECG recorded. Measurement of calibration records published by Einthoven reveals a response time (10% to 90%) of 20 ms. The corresponding sinusoidal frequency response is 0 to 25 Hz (30% attenuation). Figure HP2.8 illustrates one of Einthoven’s electromagnets (lower left) and one of his early cameras (center) and string galvanometers (right). Above the bench supporting these are mounted a bow and arrow, which were used by Einthoven to provide the tension on the quartz rod while it was being heated to the melting point, after which the arrow left the bow and extruded the rod to a long slender filament, which was later silvered to make it conducting.

© 2000 by CRC Press LLC

FIGURE HP2.5 Capillary electrometer record (dark hump) and the corrected voltage-time record (biphasic wave).

FIGURE HP2.6 A capillary electrometer record (a) and its corrected version (b) presented by Einthoven to show the inadequacy of the capillary electrometer in displaying rapidly changing waveforms. (From FA Willius, 1941. TE Keys, Cardiac Classics, St. Louis, Mosby.)

Einthoven borrowed handsomely from previous work; he used Marey’s recording chart speed (25 mm/s) and Waller’s bucket electrodes, as well as some of his leads. With the string galvanometer, Einthoven ushered in clinical electrocardiography in the early 1900s; soon the string galvanometer was used worldwide. The first string galvanometer appeared in the United States in 1912, when it was used at the Rockefeller Institute Hospital until 1959. This instrument, which is now in the Smithsonian Institute, was designed by Horatio B. Williams, professor of physiology at the College of Physicians and Surgeons at Columbia University. Williams had spent some time with Einthoven in Leiden in 1910 and 1911. On his return to New York, Williams had Charles F. Hindle, a machinist at Columbia, construct the first American string galvanometer. Soon thereafter, the Cambridge Instrument Company took over manufacture of the Hindle instrument and made them available for sale in the United States. Although it is clear that the concept of an electrical axis, i.e., a cardiac vector, was demonstrated by Waller’s studies, it remained for Einthoven [1913] to make practical use of the concept. Einthoven postulated that the heart was at the center of an equilateral triangle, the apices of which were the right and left shoulders and the point where both legs joined the trunk. In his early studies, Einthoven used the right and left arms and both feet in saline-filled buckets as the three electrodes. Soon he found that the electrocardiogram was negligibly altered if the right foot was removed from the bucket electrode. Thus he adopted three standard leads: right and left arms and left leg (foot). He postulated that if the amplitudes of the electrocardiographic waves are plotted on this triaxial reference frame, it is possible to calculate the magnitude and direction of an electric vector that produces these same voltages in leads I, II, and III, corresponding to the limb electrodes. He further stated that the arithmetic sum of the amplitudes in lead I plus III equals the amplitude in lead II. This is Einthoven’s law, and the relationship is true only for an equilateral triangle reference frame [Valentinuzzi et al., 1970].

Vacuum-Tube Electrocardiograph Not long after Einthoven described his string galvanometer, efforts were begun in the United States to create an electrocardiograph that used vacuum tubes. At that time, there were rapidly responding mirror © 2000 by CRC Press LLC

(b)

FIGURE HP2.7

© 2000 by CRC Press LLC

Einthoven’s string galvanometer (a) and a patient having his ECG (lead 1) recorded (b).

FIGURE HP2.8 Some of Einthoven’s early equipment. On the bench (left) is an electromagnet. In the center is a camera, and on the right is a string galvanometer. On the wall are a bow and arrow; the latter was used to apply force to a quartz rod which was heated, and when pulled by the arrow, created the quartz filament which was later silvered.

galvanometers, as well as a limited number of vacuum tubes, despite the fact that they has been patented only a few years earlier [1907] by DeForest. According to Marvin [1954], the first discussions relative to such an instrument were held in 1917 between Steinmetz, Neuman, and Robinson of the General Electric Engineering Laboratory. The task of establishing feasibility fell to W. R. G. Baker, who assembled a unit and demonstrated its operation to those just identified. However, because of urgent wartime priorities, the project was shelved. In 1921, General Electric reopened the issue of a vacuum-tube ECG. A second prototype was built and demonstrated to the Schenectady County Medical Association some time in 1924 by Robinson and Marvin. The instrument was used by Drs. Newman, Pardee, Mann, and Oppenheim, all physicians in New York City. Subsequently, six commercial models were made. One instrument was sent to each of the four physicians just identified; the fifth was sent to the General Electric Company Hospital; and the sixth was sent to the AMA Convention in Atlantic City in 1925. This latter instrument became a prototype for future models provided by the General Electric X-Ray Division. A U.S. patent application was filed on January 15, 1925, and the instrument was described by Mann [1930]. On December 22, 1931, a patent on the General Electric vacuum-tube ECG was granted to Marvin and Leibing; Fig. HP2.9 shows the circuit diagram of the instrument, including a specimen record (lower left, Fig. HP2.9). The instrument used three triode vacuum tubes in a single-sided, resistance-capacitancecoupled amplifier. It was battery operated, and a unique feature was a viewing screen that allowed the operator to see the motion of the galvanometer beam as it was being recorded by the camera. Between introduction of the string galvanometer and the hot-stylus recorder for ECG, attempts were made to create direct-inking ECG recorders. In a review of scientific instruments, Brian Matthews [1935] reported © 2000 by CRC Press LLC

FIGURE HP2.9

Circuit diagram of the first vacuum-tube ECG, patented on December 12, 1931.

There are two ink-writing electrocardiographs available, the author’s and that of Drs. Duschel and Luthi. Both utilize a moving-iron driving unit with oil damping, a tubular pen writing on moving paper. The former has a battery-coupled amplifier. The latter gives, in effect, D.C. amplification; the potentials to be recorded are interrupted about 500 times per second by a special type of buzzer, after © 2000 by CRC Press LLC

amplification by resistance capacity coupled valves the interrupted output is reflected by the output valve; the amplifier achieves in effect what can be done with a battery-coupled amplifier and obviates the coupling batteries. At present the speed of these direct-recording instruments is barely adequate to show the finer details of the electrocardiogram, but they enable its main features to be recorded instantly. Despite the instant availability of inked recordings of the ECG, those produced by the string galvanometer were superior, and it took some time for a competitor to appear. Such an instrument did appear in the form of the hot-stylus recorder.

Hot-Stylus Recorder The final step toward modern electrocardiography was the introduction of the hot-stylus recorder by Haynes [1936] of the Bell Telephone Laboratories (New York). Prior to that time, there was colored, waxcoated recording paper, the wax being scraped off by a recording stylus, exposing the colored paper. Referring to the scraping method, Haynes wrote However, the method is not adaptable to many types of recording instruments because of the large amount of friction between the recording stylus and paper arising from the pressure necessary to engrave the wax. This pressure can be removed and the friction largely eliminated by the use of a special stylus consisting essentially of a small electric heating coil situated close to the end of a pointed rod in such a way that the temperature of the point may be raised to about 80°C. The point is then capable of melting wax and engraving its surface with only a very small fraction of the pressure before necessary. The described stylus, when used with waxed recording paper, will provide a means of obtaining a permanent record without the use of pen and ink which is adaptable to the most sensitive recording instruments. Following the end of World War II, vacuum-tube electrocardiographs with heated-stylus recorders became very popular; they are still in use today. However, the heritage of a thick baseline, derived from the string-galvanometer days, had to be preserved for some time because clinicians objected to a thin baseline. It took many years for the hot-stylus baseline to be narrowed without protest.

References Ader M. 1987. Sur un nouvel appareil enregistreur pour cables sousmarins. C R Acad Sci 124:1440. Burch GJ. 1892. On the time relations of the excursions of the capillary electrometer. Philos Trans R Soc (Lond) 83A:81. Burch GJ. 1892. On a method of determining the value of rapid variations of potential by means of the capillary electrometer communicated by J. B. Sanderson. Proc R Soc (Lond) 48:89. Burdon-Sanderson JS, Page FJM. 1879. On the time relations of the excitatory process of the ventricle of the heart of the frog. J Physiol (Lond) 2:384. Caton R. 1875. The electric currents of the brain. Br Med 2:278. DeForest L. U.S. patents 841,387 (1907) and 879,532 (1908). Donders FC. 1872. De secondaire contracties order den involed der systolen van het hart, met en zonder vagus-prikkung. Oncerszoek ged in physiol Lab d Utrecht Hoogesch, Bd. 1, S. 256, Derde reeks TI p 246 bis 255. Einthoven W. 1903. Ein neues Galvanometer. Ann Phys 12 (suppl 4):1059. Einthoven W, Fahr G, de Waart A. 1913. Uber die Richtung und die manifeste Grosse der Potential schwankunzen in menschlichen Herzen. Pflugers Arch 150:275. Garceau EL, Davis H. 1935. An Ink-writing electroencephalograph. Arch Neurol Psychiatry 34:1292.

© 2000 by CRC Press LLC

Gasser HS, Erlanger J. 1922. A study of the action currents of nerve with the cathode ray oscillograph. Am Physiol 62:496. Gasser HS, Newcomer HS. 1921. Physiological action currents in the phrenic nerve. Am Physiol 57(1):1. Geddes LA, Foster KS, Senior J, Kahfeld A. 1989. The Inductorium: The stimulator associated with discovery. Med Instrum 23(4):308. Haynes JR. 1936. Heated stylus for use with waxed recording paper. Rev Sci Instrum 7:108. Hoff HE, Geddes LA. 1957. The rheotome and its prehistory: A study in the historical interrelation of electrophysiology and electromechanics. Bull Hist Med 31(3):212. Kolliker RA, Muller J. 1856. Nachweiss der negativen Schwankung des Muskelstromsnaturlich sic contrakinenden Muskel: Verhandl. Phys Med Ges Wurzburg 6:528. Mann H. 1930–31. A light weight portable EKG. Am Heart J 7:796. Marey EJ. 1885. Methode Graphique, 2d ed. Paris, Masosn. Marey EJ. 1876. Des variations electriques des muscles du coeur en particulier etudiee au moyen de l’ectrometre d M. Lippmann. C R Acad Sci 82:975. Marvin HB, et al. 1925. U.S. patent 1,817,913. Matthews BHC. 1935. Recent developments in electrical instruments for biological and medical purposes. J Sci Instrum 12(7):209. Valentinuzzi ME, Geddes LA, Hoff HE, Bourland JD. 1970. Properties of the 30° hexaxial (EinthovenGoldberger) system of vectorcardiography. Cardiovasc Res Cent Bull 9(2):64. Waller AD. 1887. A demonstration on man of electromotive changes accompanying the heart’s beat. J Physiol (Lond) 8:229.

© 2000 by CRC Press LLC