Gordon E. Dower

Deriving the 12-lead electrocardiogram from four (EASI) electrodes☆

Computerized interpretation of the electrocardiogram has now advanced to computerization of the electrocardiograph, resulting in greatly increased versatility, including the capacity for adapting to a variety of lead systems rather than being tethered to the old Einthoven-Wilson-Goldberger (EWG) system. Many varieties of display beyond the 12-lead ECG are also available in software. To date, these new and interesting capabilities have scarcely been exploited. The EASI lead system uses the E, A, and I electrode positions of the Frank lead system, plus an electrode, S, positioned over the upper end of the sternum and, if necessary, ground (anywhere convenient). Its outputs form quasi-xyz signals, xyz, that can be approximately transformed into xyz signals by means of a matrix derived from the EASI lead vectors. The result forms a good basis for deriving the 12-lead ECG, using previously published coefficients for the Frank lead system. The match with the conventional ECG can then be improved by statistical means. The results are surprisingly good, and certainly of clinical value. Recent widespread interest in silent ischemia and its detection through Holter monitoring suggests an immediate application which has been rendered practical by the recent introduction of three-channel recorders. The EASI electrode positions give technically satisfactory Holter recordings. Very compact three-channel, multiplexed, radio telemetry equipment is now commercially available and provides another application for the EASI 12-lead ECG.(ABSTRACT TRUNCATED AT 250 WORDS)

THE POLARCARDIOGRAPH. TERMINOLOGY AND NORMAL FINDINGS.

Abstract The clinical application of the Polarcadiograph has brought out the need for a comprehensive, yet concise, terminology to describe the polar coordinates of the heart vector. In the terminology proposed, precedent has been followed insofar as consistency and clarity permit. Frontal, transverse, and sagittal longitudes, designated by α, β, and γ, have been based on Einthovens precedent for α. The corresponding latitudes are indicated by P-A, I-S, and R-L. The spatial magnitude is indicated by M, and the magnitudes in the frontal, transverse, and sagittal planes by fm, tm, and sm. By integration of the spatial magnitude tracing produced by the QRS vectors a point is found at which the area under this curve is half inscribed. The QRS vector associated with this point in time is termed IR. It is suggested that this be considered representative of the QRS vectors. Other defined vectors are R, the peak QRS vector, and T, the peak T vector. The usefulness of the Polarcardiograph derives from the fact that it provides a fresh display of the electrical events associated with the cardiac cycle. P-R, QRS, and Q-T intervals are clearly indicated by the spatial magnitude tracing and have their usual significance. However, difficulties arising out of asynchronous isoelectric portions of the electrocardiogram are avoided. In a group of 192 elderly men with completely normal electrocardiograms, the QRS interval determined from the spatial magnitude tracing was found to be between 61 and 105 msec. for 95 per cent of them. The distributions of values for the spherical polar coordinates of R and T are given. The relationship between R and T was indicated by the distribution of their magnitude ratio, MTMR, and of the subtended angle. The spatial magnitude of R, MR, was between 0.7 and 2.0 mv. in 95 per cent of the cases. (A small sample of athletes showed much larger values.) The spatial magnitude of T, MT, was between 0.19 and 0.74 mv. The directions of R covered approximately one third of the globes surface. Most remarkable was the relatively small scatter of the directions of T, of which 95 per cent were contained in one fifteenth of the globes surface.

THE POLARCARDIOGRAPH. DIAGNOSIS OF MYOCARDIAL INFARCTION.

Abstract The usefulness of five simple criteria for the polarcardiographic diagnosis of myocardial infarction was evaluated in 493 patients, of whom 227 were considered, on the basis of ECG or postmortem evidence, to have suffered infarction. In these, the criteria gave the diagnosis in 94.7 per cent. On the other hand, two independent readers, using conventional criteria, reached the diagnosis from the single ECG, recorded at the same time as the PCG, in only 66.1 per cent. In 11.6 per cent of the total population the criteria indicated myocardial infarctions which could not be confirmed by available evidence. The results indicate that the Polarcardiograph may be of considerable value in the diagnosis of coronary heart disease.

The Polarcardiograph. Further studies of normal subjects, refinement of criteria for infarction, and a report on autopsied cases☆

Abstract A population sample of 195 young men and women has been studied in order to define normal values for polarcardiographic diagnosis and to refine five simple criteria for infarction to yield an incidence of false positives of less than 5 per cent. The incidence became: positives, 3 per cent; doubtfuls, 2 per cent. The criteria were reapplied to a group of 493 patients who had been studied previously, and to 78 autopsied cases which formed a subgroup. In the 493 patients there were 234 who were considered to have suffered infarction. In these, the refined criteria gave the diagnosis in 88 per cent. On the other hand, two independent readers, using conventional criteria, reached the diagnosis from the single ECG, recorded at the same time as the PCG, in only 64 per cent. In the 78 autopsied cases there were 58 with infarcts. The ECG gave the diagnosis in 40 per cent of these, whereas the PCG gave the diagnosis in 79 per cent.

A clinical comparison of three VCG lead systems using resistance-combining networks

Abstract Two 4-electrode, VCG lead systems have been compared, under clinical conditions, with a theoretically more accurate 7-electrode system, in 40 patients. All systems used the appropriate resistance-combining networks embodying correcting factors determined from torso models. One of the 4-electrode systems, the RLFB, which was based on the Wilson tetrahedron, had been compared with the 7-electrode system by Frank and Seiden, and had been rejected on the grounds of insufficient accuracy. The other 4-electrode system, the RAFE, was designed, unlike previous systems, to give optimal convenience of application. The results showed: 1. 1. Under conditions obtaining with ill patients, the RAFE showed a highly significant advantage with respect to accuracy as compared with the RLFB. This was due largely to improvement in the recording of sagittal components. 2. 2. The signal-to-noise ratio with the RAFE was close to that obtained with the 7-electrode system and much greater than that with the RLFB. The results indicate that the optimal convenience provided by the RAFE lead system is achieved at a relatively small cost in accuracy and signal magnitude. They suggest the use of this electrode arrangement at least in seriously ill patients, who would be unduly disturbed by the application of the 7-electrode system.

On electrocardiographic-autopsy correlations in left ventricular hypertrophy. A simple postmortem index of hypertrophy proposed☆

Abstract Review of the literature has brought out the following: (1) ECG criteria for LVH commonly give false positives and false negatives. (2) Although LAD, prolonged VAT and ST-T changes may be of value, the most useful are the voltage criteria, of which there are many. (3) Low correlation between increased voltage and increased heart weight at autopsy may derive from technical errors in ECG recording that arise from the equipment used and the use of criteria of normality which may exclude heart weights which are actually normal. (4) Examination of postmortem criteria of normality reveals that a satisfactory definition of LVH does not exist. A simple index of hypertrophy based on a reanalysis of data on 926 normal cases is proposed: HW BL > 2.20 Gm. per centimeter, for men, and > 2.06 Gm. per centimeter for women.

The polarcardiograph. Diagnosis of left ventricular hypertrophy

T he Polarcardiograph gives the direction and magnitude of the heart vector as continuous tracings known as polarcardiograms (PCGs).’ In the diagnosis of infarction the superiority of the PCC over the single electrocardiogram (EC<;), recorded at the same time, has been demonstratedc2 in a group of autopsied cases the PCG was found to be almost twice as sensitive an indicator of infarction.3 In the ECG the large QRS amplitudes associated with left ventricular hypertrophy (LVH) are manifestations of an increase in the spatial magnitude of the heart vector. Gamboa and associates4 have shown that the magnitude of the maximum heart vector correlates strongly with maximum systolic pressure in the left ventricle and hence with various forms of pressure overload. Since the PCG displays the magnitude of the heart vector as a time graph, it should provide a simple indication of LVH. What is LVH? It seems almost axiomatic that the most rigorous definition of LVH would be made on anatomic grounds. Perhaps the most widely used postmortem (PkI) criteria of LVH are those based on upper bounds for heart weight for various body lengths reported by Zeek” for a large series of normal men and women. Reexamination of Zeek’s raw data by Dower and associates,‘j while it confirmed her finding of a significant correlation between heart weight and body length, disclosed that her table excluded 27 per cent of her normal men and 47 per cent of her normal women. Use of these criteria would therefore give rise to a high incidence of false positives in the PM diagnosis of LVH. Reanalysis of Zeek’s data yielded a simple index of hypertrophy, the heart-weight: body-length ratio (HW/BL). The upper j-percentile points w:ere 2.20 (ini. per centimeter for men and 2.06 (;m. per centimeter for women (Table I). These figures will be used as an anatomic definition of LVH in this paper. The diagnosis of LVH from the EC(; hinges largely upon the so-called voltage criteria.” However, the popular choice of ECG criteria for LVIH seems to pose somewhat of a problem. Dower and associatesG reviewed 17 studies of the correlation between the diagnosis of L:H on ECG and PRI grounds which disclosed that 43 different voltage criteria were used, although the differences between some of these were

Editorial Some Instrumental Errors in Electrocardiography

IT IS perhaps not surprising that the invention of the first true electrocardiograph should be associated with the first statement concerning the response characteristics necessary to obtain records having minimal or negligible distortion. In 1912, Einthoven expressed the view that if, following the application of a constant potential difference, the galvanometer reached its new position of equilibrium within about 0.01 second, distortion of the electrocardiogram would be negligible.1 This corresponds to the rise-time, or deflection-time in modern usage, which is generally measured between the 10 and 90 per cent points of the deflection curve. An instrument having a satisfactory rise-time may yet yield distorted records if it is underdamped. Ideally, it should be critically damped, but a slight degree of underdamping is often sought as a compromise between the avoidance of resonance, on the one hand, and the preservation of a sufficiently rapid response on the other. In their recommendations concerning the minimum requirements for acceptable electrocardiographs, the Council on Physical Medicine and Rehabilitation of the American Medical Association, in 1950, stipulated that the amplitude response of the instrument to 1 millivolt peak sinusoidal voltage variation up to 300 cycles per second shall not exceed

A simple test of speed of response of electrocardiographs

Abstract A convenient test of the speed of response of an electrocardiograph can be provided by a very simple circuit. This provides an exponential pulse with a time-constant of 6.3 milliseconds and an equal voltage step. For a direct-writer which barely meets the minimum requirements of the American Medical Association the recorded amplitude of the pulse is approximately 50 per cent that of the step. If the tester is incorporated in the conventional calibration circuit, only a capacitor and a button switch need be added.

MEASUREMENT OF THE ERROR IN WILSON'S CENTRAL TERMINAL: AN ACCURATE DEFINITION OF UNIPOLAR LEADS

Unipolar leads using Wilsons central terminal as an indifferent electrode have been in general clinical use for many years. The central terminal was not introduced as a theoretically correct reference for unipolar leads, but as a convenient reference that was at least better than using one of the limbs alone, as in the old CR, CL, and CF leads, in which the exploring electrode potentials were compared with those of electrodes placed respectively on the right arm, the left arm, or the left leg (Wilson et al., 1934). When unipolar leads are used to explore the potentials over the left chest at the level of the heart, the error in adopting the central terminal as a reference electrode is small because the potential variations of the exploring electrode are relatively large. A tendency to explore the right chest has developed in recent years. In this case, the error may be considerable because the smaller potential variations of the exploring electrode may be ofa magnitude comparable with those of the central terminal itself. Such records may be open to misinterpretation. A case has recently been described in which the unipolar lead tracings V3R were produced by potential variations of the central terminal rather than by those of the exploring electrode (Osborne and Dower, 1957), and resulted in a pattern resembling right ventricular hypertrophy for this lead. Methods of determining the error in the central terminal have been described by Eckey and Fr6hlich (1938), Burger (1939), Wilson et al. (1949), Dolgin et al. (1949), Bayley and Kinard (1954), and Frank (1955a), but none has been applicable to patients. For this reason, it has not been possible to study the error under clinical conditions. The method to be described makes this possible. The theoretical considerations that arose out of the development of this method have demanded a careful scrutiny of previous definitions of unipolar leads and this has led unexpectedly to the conclusion that a more accurate definition is required.