The First Artificial Pacemaker, 1928-32 In the second of the three papers on "intracardial therapy" (1932), Hyman described and illustrated a device for physician use, for which he and his brother Charles received a patent in 1933. The device, which emitted electrical stimuli in order to restart the "stopped heart" and which he named the "artificial pacemaker" (first use of that term), which since has achieved widespread acceptance.
In addition to announcing the invention, Hyman went to some lengths to provide an acceptable rationale for electrical cardiac stimulation. He thought of the electrical stimulus as little more than a substitute for the needle-prick. "Of fundamental importance," he wrote, "is the difference in the basic theories between the previous modes of electrically stimulating the heart and that concerned with artificial pacemaker methods. In the former, the electric current introduced into the heart is the same current that is supposed to cause contraction of the heart muscle tissue, whereas in the latter theory the introduced electric impulse serves no other purpose than to provide a controllable irritable point from which a wave of excitation may arise normally and sweep over the heart along its accustomed pathways."
"In other words," he went on, "the artificial pacemaker produces the same effect as that previously discussed in regard to the mechanical prick of an injecting needle"
Hyman had actually begun to design his pacemaker in April 1928 with the assistance of his brother Charles H. Hyman, a self-described physicist. Then, in September 1929, Australian physician Mark C. Lidwill announced before the Australasian Medical Congress that he had revived at least one stillborn infant by means of an electrostimulation device. However, Lidwill’s brief discussion (reprinted by Mond et al. in 1982) provided no technical details of his invention. The Hymans’ patent application for their "Artificial Pace Maker for the Heart" is dated 12 March 1930; they had completed a working prototype of their invention by early 1931.
During 1931-32, Albert Hyman employed the artificial pacemaker on small laboratory animals (rabbits, guinea pigs, and "one large dog") that had been brought to cardiac standstill through asphyxiation or other means. The pacemaker delivered ventricular stimuli because Hyman found it difficult if not impossible to direct the needle electrode to the atrium. Hyman claimed that with the pacemaker he had restored the heartbeat in a number of animals; many years later, he stated that the invention had revived 14 out of 43 animals. In his 1932 paper, Hyman claimed that the artificial pacemaker stimuli were shown by ECG to produce "extrasystoles." But review and reanalysis of those published ECGs does not persuade us that an actual cardiac response had occurred. (Figure 1)
| |
Figure 1: Figure 10 of Hyman’s 1932 publication (shown below) purported to demonstrate the ability of the pacemaker to stimulate the right ventricle. A) A prolonged period of asystole caused by asphyxiation of a guinea pig. B) Pacemaker stimulation stated to produce ventricular contractions. While the stimuli are clearly present, the documented response is only a return to baseline without an indication of a QRS complex. After return to baseline, there is no T wave, the unambiguous indicator of a preceding ventricular depolarization. C) Recovery results in return of a slow spontaneous rhythm with P, QRS and T waves. D) Return of a normal, spontaneous rhythm. This group of ECGs does not demonstrate that a ventricular response followed the pacemaker stimuli. 
Fig. 10 (in the original). Resuscitation of the asystolic heart. A large guinea-pig was mechanically asphyxiated. A, electrocardiographic tracing taken in lead II, during the period of cardiac standstill. B, about forty seconds later, undampened pacemaker stimuli were released in the right ventricle at the rate 120 per minute. Note the bizarre type of right ventricular extrasystoles that developed. C, removal of the pacemaker needle was followed by a normal sinus rhythm corresponding in rate with that established by the artificial pacemaker. D, restoration to regular normal sinus rate and rhythm, the pacemaker of the heart now having taken over control of the cardiac cycle.
| |
How it Worked
Three different models of the Hyman artificial pacemaker were built in the 1930s, but none survives today. A few published photographs remain (Figures 2-5), and a separate photo(Figure 6) reveals the more finished-looking device named the Hymanotor by its manufacturer . We know of no photograph of the third or "flashlight" version, to which he referred only in a newspaper interview. In 1998 we located the United States patent for the first Hyman pacemaker and one of us (Szarka) constructed a partially working replica.
| |
Figure 2: A computer enhancement of Figure 1 of the 1932 Hyman manuscript (shown below) shows additional detail: C) the neon lamp which is illuminated when a stimulus is withheld; D) the spring motor with winding gear and teeth, now clearly visualized; E) a ballistic governor speed control with three centrifugal weights, for the spring motor; F) handle with ratchet to permit rewinding while the spring motor is unwinding, i.e., in motion; G) impulse controller to regulate position of the interrupter brushes to give the three positions, 30, 60, 120 impulse per minute; H) speed control for the spring motor which can also start and stop the apparatus; I) wire connecting the pacemaker to the needle assembly; J) the needle handle; K) a switch on the needle handle; N) possibly the two connector terminals for the wire to the needle; X) slot head wood screw. 
Fig. 1 (in the original). The artificial pacemaker seen from the front. In figures 1 to 4 the following important features are to be noted: A, magnetogenerator; B 1 and B 11, companion magnet pieces; C, neon lamps; D, spring motor; E, ballistic governor; F, handle; G, impulse control; H, speed control; I, flexible electric cord; J, insulated handle; K, handle switch, and L, electrode needle. | |
| |
Figure 3: A second illustration from 1932 (shown below) reveals additional features: M) one face of the interrupter disk; X) wood screws; V) coiled wire; N(?) the terminals to connect pacemaker output to the needle cable.
Fig. 2 (in the original). The artificial pacemaker seen from the back. The interrupter disk is shown at M. | |
| |
Figure 4: In their patent narrative, the Hymans stated that "the needle consists of two electrodes, one an outer tubular sleeve . . . and the other a metal core surrounded by an insulating sheath." In this photograph from 1932, L) is a hollow #19 gage steel shaft through which an insulated wire passes and terminates at the same angle as the outside needle shaft; R and S) are the electrical connections; R the needle shaft and S the inside (wire) electrode; M) is a tube holder for six glass tubes to hold sterile needles; P) is a tube; Q) is an impermeable stopper.
Fig. 3 (in the original). Photograph showing details of needle (L), its electric connections (R, S), handle (J) and switch (K). The special tube holder ( M 1) and tube (P) with stopper (Q) are also seen. | |
| |
| Figure 5: This exterior view shows O) the instruction card; T) the cover which can be held down by Y; Y) trunk catches; Z) a chain to hold the lid in place; M1) the tube holder with six stoppers seen (Q); N) one of two terminals to connect to the needle. The pacemaker is its carrying case. The hand crank may have been stored inside the lid and is not seen in the photograph. The entire assembly weighed 16 lbs. (7.2 Kg). | 
Fig. 4 (in the original). The artificial pacemaker in its carrying case. | | |
| |
| 
Figure 6: The Hymanotor, a new model of Hyman’s pacemaker probably intended for laboratory experiments with animals, 1933. Several features are evident: the needle electrode (in front of the pacemaker box); an empty pair of clips to hold the electrode handle (in the lid); glass tubes containing sterile needles ready for use (also in the lid); a hand crank for winding the spring motor (stored in the lid at the left) and a socket for the crank (right side of the housing); a voltage control knob (front left); a rate control mechanism with a thumb screw or button for switching between 30, 60, and 120 impulses per minute (just above the "Hymanotor" name plate); an on-off switch (above the voltage knob); a small neon lamp that flashed to indicate delivery of stimuli (to the right of the speed control); two knobs around which the electrode cord could be wound (back corners of the console). The Hymanotor had two sockets into which the electrode wire could be plugged, one labeled "Stimulation,"the other "Needle Test" (on either side of the name plate). The meter at the center of the console apparently gave voltage readings when the "Meter" button (to the left of the meter) was depressed. It is possible that the operator had to depress the right-hand button, labeled "Impulses," for the machine to deliver stimuli to the heart. The function of the knob at lower right on the console could not be determined. Photo courtesy of the Bakken Library and Museum, Minneapolis, MN, USA. |
| |
To generate a periodic pacing waveform, the Hyman brothers employed a magneto-generator driven by a spring motor and gear train. A magneto is a generator of a direct current voltage; it was widely used at that time to start automobiles and to generate a bell ringing voltage for telephones. The large "U" shaped magnets on the original Hyman pacemaker, often mistaken for handles, provided the magnetic flux necessary to generate current in the magneto-generator. As the conductors in the generator windings cut through the magnetic flux field, a current was induced in the conductors. This current was gated through the interrupter disc to drive lamps and produce pacing stimuli.
In the original model, the output voltage of the magneto-generator (and hence, the current level) was controllable by adjusting the speed at which it was driven. (Figure 7) The hand crank wound the spring motor, which was then able to deliver the rotational power to the magneto-generator. In a drive mechanism similar to that found in a wind-up music box, a mechanical ballistic governor, adjustable by means of a thumb-wheel screw accessible at the top of the device, regulated the motor speed. The operator had only to wind the pacemaker to store rotational energy. Continuous hand cranking was not required to generate a succession of stimuli.
| |
Figure 7: The complexities of the pacemaker yielded a simple operational concept. The hand crank winds the spring motor which drives the magneto-generator at a governor controlled speed and causes the interrupter disc to rotate. The magneto-generator supplies current to a surface contact (referred to as a brush) which in turn makes intermittent contact with the rotating conductive surfaces on the interrupter disc, separated by an insulated surface. Each time such contact is made the illuminated neon lamp is extinguished and an electrical surge passes to the needle and then to the heart. 
| |
In addition to driving the magneto-generator, the spring motor also propelled the interrupter disc. The double-sided interrupter disc consisted, on one side, of four conductors of different lengths, radiating at right angles (i.e. 90° ) from the disc center. The remaining area on that side of the disc served as an electrical insulator. Two brushes or contacts wiped the disc surface, providing a path for current flow when the contacts passed over the conductor. These contacts delivered the stimulation voltage to the needle electrode. The obverse of the disc provided an identical pattern, but with the conductive areas and insulating areas reversed. A similar set of contacts existed on this surface, providing voltage to indicator lamps which illuminated when the stimulus voltage was interrupted and extinguished when the stimulus voltage flowed.
Problems with the Invention
Careful examination of the patent suggests that the original Hyman pacemaker design had problems requiring further development work. As the conductive tracks on the "shock" side of the interrupter disc radiated from the center at right angles, this pattern yielded four distinct stimulus pulse rates in multiples of 1 to 4 times the revolutions-per-second of the interrupter disc. At 1, 2 and 4 stimuli per disc revolution, stimuli were regularly generated at 30, 60 and 120 per minute. At three stimuli per revolution the stimulation rate is 90 asymmetrically spaced pulses per minute, irregularly spaced. Perhaps for this reason Hyman did not claim this rate for the pacemaker. Szarka’s proposed revision regularizes all four rates . Machines are usually designed so that the center of an adjustment range is reserved for the "nominal" setting, so it would seem that this peculiarity occurred at the worst possible spot. (Figure 8)
Another problem with the Hymans’ design resulted from driving the magneto-generator and interrupter disc together, by the same source, the speed-regulated spring motor. Adjusting the speed of the motor simultaneously affected pulse rate, stimulus voltage, and current level, an undesirable association which could have been avoided by the addition of a second speed regulator. The patent narrative did not indicate the pacemaker’s pulse duration, our analysis of this important parameter required assumptions:
- Diameter of Interrupter Disc = 3 inches, Radius = 1.5 inches
- Base speed of the Interrupter Disc = 0.5 revolutions/s
- Width of conductive track at outer edge of disc = 0.25 inch
| |
 Figure 8: (A) The original interrupter disc is clearly illustrated in figure 6 of the Hyman brothers’ patent. Generating pulses with the illustrated pattern compels pulse duration to be linked to the pulse repetition rate. The pulse duration is proportional to the to the number of impulses per minute, i.e. the stimulation rate. At a higher stimulation rate the pulse duration is of lesser duration. (B) The Szarka correction of the interrupter disc, though never built, would have been capable of working with the speed control mechanism. In this formulation the speed control brush would continue to move vertically from adjacent to the axis to the outer rim of the disc. Four positions are possible; the position closest to the axis would contact the conductor once per rotation (rate 30 stimulus per minute); one step out, contact would be made with two conductors (rate 60 stimuli per minute, at equal intervals); one step further, contact occurs with three contacts (rate 90 stimuli per minute, at equal intervals) and with the brush closest to the rim, contact occurs with four contacts (rate 120 stimuli per minute, at equal intervals). This sequence is reversed from that designed by the Hymans, in which the most rapid rate occurred with the brush nearest the axis, rather than the disc rim. (C) The stimulation rate control assembly (G) holds the brushes in contact with the interrupter disc, shown vertically. Knob rotation of the threaded shaft moves the brushes vertically between the upper and lower stops, disc center to rim. |
| |
Based on the original conductive 4-track pattern of the surface of the interrupter disc and with the base speed of the interrupter disc set at 0.5 revolutions per second, the lowest stimulation rate obtainable would be 30 BPM. The maximum rate would be 120 BPM. (Figure 9)
Using 2p *radius, the circumference of the disc is 9.42 inches and the angular velocity at the disc edge is 9.42 inches/2 seconds (note that a revolution takes 2 seconds) or 4.71 inches per second. Assuming an angular velocity of 4.71 inches per second and a track width of 0.25 inches, conduction would occur for 4.71 in./s* 0.25 in. = for 53 milliseconds. Based on these assumptions, we believe that the stimulus pulses would have had a duration of approximately 50 milliseconds.
| |
Figure 9: This is a diagrammatic representation of the pattern of stimulation at the various rate settings of the interrupter disc in the Hyman design. It is not implied that the stimuli were rectangular pulses. At one pulse/rotation the rate would have been 30/minute; at two pulses, 60/ minute; at 3 pulses an irregular rate of 90/minute and at 4 pulses; 120/minute. With the Szarka revision, all stimuli would have been evenly timed.
| |
;)
;)
;)
;)
;)
;)
;)
;)
;)
