Quantitation of exercise electrocardiography

Clinical evaluation of algorithms for ST measurement during exercise test

HYPOTHESIS

Computer processing of the exercise electrocardiogram (ECG) has many advantages, but the reliability of the analysis algorithms is not easily evaluable. No standard annotated database, nor recommended practice for testing and reporting performance results is available: thus, performance evaluation of such devices can be accomplished only by using a set of unannotated recordings, obtained in clinical practice. We evaluated the accuracy of an original microcomputer-based exercise test analyzer comparing the ST computer output with the measurements obtained by two experienced cardiologists.

METHODS

Six hundred ECG strips were randomly selected from the exercise test recordings of 60 patients. The ST shift (at J + 80 ms) was blindly assessed by two observers (with the aid of a calibrated lens) and compared with computer measurements. Correlation coefficients, linear regression equations, percent of discrepant measurements, and 95% confidence limits of the mean error were calculated for all leads, peripheral leads, precordial leads, and "stress-test" leads (II, III, aVF, V4, V5, V6).

RESULTS

The computer did not analyze five samples on a total of 600 (0.83%) ECG strips because of excessive noise or signal loss, while 51 (8.5%) were considered unreadable by both observers and 67 (11.2%) were rejected by at least one observer. Correlation between the measurements taken by computer and observer(s) measurements was statistically significant (p < 0.001 for all lead groups), no systematic measurement bias was found, and the mean difference was lower than human eye resolution.

CONCLUSIONS

Our algorithms provide results as good as those provided by trained cardiologists in measuring ST changes occurring during exercise test. However, this study did not evaluate whether computer improvement of the signal-to-noise ratio would allow accurate measurements even on cardiologists' uninterpretable ECG. This potential advantage of computer-assisted analysis could be assessed only by using a dedicated exercise test database, in which different patterns of noise are superimposed on noise-free recordings previously annotated for ST level.

Compartmental multivariate analysis of exercise ECGs for accurate detection of myocardial ischaemia

An accurate computer-assisted diagnostic method for detection of myocardial ischaemia, called MUSTA, is developed. MUSTA is based on compartmental multivariate analysis of variables available in the exercise ECGs, and is definitively implemented in Prolog. It is heuristically developed by determining diagnostic criteria, which interrelate a modified ST/HR-slope, ST-segment value and shape, and maximum heart rate, so that concordance with the TI-201 SPECT is maximised. In the learning group consisting of 47 patients, MUSTA provides a diagnostic accuracy of 98%, the detection of ischaemia being in absolute concordance with TI-201 SPECT. MUSTA is evaluated in a similar but independent group of 60 patients. Then, accuracy is 90%, and sensitivity is 94%. The performance characteristics are significantly better than those of the standard exercise ECG, whose diagnostic accuracy in these groups is 77% and 70%, respectively. This study suggests that MUSTA is a significant improvement for computerised assessment of myocardial ischaemia.

Reproducibility of computer measurement of maximal isometric strength and electromyography in sedentary middle-aged women

The objectives of this study were (i) to determine the reproducibility of computer measurements of isometric strength and related electromyography in several muscle groups in sedentary middle-aged women, (ii) to evaluate the effects of different digital signal averaging methods on the reproducibility, (iii) to determine the final test score to be preferred in terms of improved reproducibility of isometric strength measurements, and (iv) to evaluate potential advantages provided by the computer measurement. Fifteen subjects were measured three times within a 2-week period. The measurements consisted of recordings of maximal isometric strength and rate of force production during trunk extension and flexion, leg extension and dominant forearm flexion with simultaneous recordings of surface electromyography, except in the trunk flexors. The following four final test scores were determined for each trial: the maximum of the three scores, the mean of the two highest scores, the median of the three scores and the mean of the three scores. The scores for the strength measurement were generally more reproducible (coefficient of variation, CV, approximately 6% and intraclass correlation coefficient ICCC, approximately 0.90) than those of the other measurements (CV > 10%, ICCC 0.13-0.97). There was no obvious preference for any type of final test score or for the width of the averaging window in the computer analysis. For isometric strength the reproducibility of the computer measurements was comparable to that of the voltmeter assessments. Computer analysis seems to be a versatile method for determining parameters of neuromuscular performance with reasonable reproducibility.

Application of computerized exercise ECG digitization. Interpretation in large clinical trials

The authors report on a semiautomated program that incorporates both visual identification of fiducial points and digital determination of the ST-segment at 60 ms and 80 ms from the J point, ST slope, changes in R wave, and baseline drift. The off-line program can enhance the accuracy of detecting electrocardiographic (ECG) changes, as well as reproducibility of the exercise and postexercise ECG, as a marker of myocardial ischemia. The analysis program is written in Microsoft QuickBASIC 2.0 for an IBM personal computer interfaced to a Summagraphics mm1201 microgrid II digitizer. The program consists of the following components: (1) alphanumeric data entry, (2) ECG wave form digitization, (2) calculation of test results, (4) physician overread, and (5) editor function for remeasurements. This computerized exercise ECG digitization-interpretation program is accurate and reproducible for the quantitative assessment of ST changes and requires minimal time allotment for physician overread. The program is suitable for analysis and interpretation of large volumes of exercise tests in multicenter clinical trials and is currently utilized in the TIMI II, TIMI III, and BARI studies sponsored by the National Institutes of Health.

Exercise-induced ST depression and ST/heart rate index to predict triple-vessel or left main coronary disease: a multicenter analysis

The aim of this investigation was to determine the difference in accuracy between two frequently published noninvasive indicators of severity of coronary artery disease (exercise-induced ST segment depression and heart rate-adjusted ST depression [ST/HR index]). The study was designed as a survey of consecutive patients undergoing exercise electrocardiography and coronary angiography. There were a total of 2,270 patients without prior myocardial infarction or cardiac valvular disease referred for angiography from eight institutions in three countries; 401 of these patients had triple-vessel or left main coronary artery disease. The sensitivities of ST depression and ST/HR index in detecting triple-vessel or left main coronary artery disease were, respectively, 75% and 78% (p = 0.08) at cut point values where their specificities were equal (64%). This small increase in the accuracy of the ST/HR index was evident only at peak exercise heart rates below the median value of 132 beats/min, where the sensitivities of ST depression and ST/HR index were 73% and 76% (p = 0.03), respectively, at cut point values corresponding to a specificity of 60%. These results were consistent at all eight participating institutions. The increase in accuracy achieved by dividing exercise-induced ST depression by heart rate is small and confined exclusively to a low exercise heart rate. This lack of superiority cannot be generalized to all methods of heart rate adjustment.

Exercise testing: uses and limitations considering recent studies

Accurate use and interpretation of exercise test results depend on an understanding of physiologic principles, meticulous attention to proper methodology, and realization of the appropriate applications and limitations of testing. Understanding the relationship between myocardial and ventilatory oxygen consumption and exercise test variables will aid in the diagnosis and prognostic evaluation. Use of proper methodology in preparing the patient, performing the examination, and interpreting the results is critical to obtaining the maximum information with maximum safety for each individual patient. Improvements in methodology including the use of the Borg scale to estimate individual effort, abandonment of the predicted maximum heart rate, and the increased use of ventilatory oxygen uptake measurements should be applied. Exercise capacity should not be reported in total time but rather as the VO2 or MET equivalent of the workload achieved. This permits the comparison of the results of many different exercise testing protocols. The most useful exercise ECG variable for the diagnosis of coronary artery disease remains the ST segment shift. Unfortunately, it is not as helpful in localizing myocardial ischemia. Diagnostic accuracy can be improved by adjusting ST depressions for exercise-induced heart rate increase. Accuracy can be further increased by combining ECG, clinical, and radionuclide variables in probabilistic formulas that retain the independent diagnostic information from each variable and accurately predict disease probability. To avoid errors in clinical decision making, care must be used to insure that the mathematical formula used was derived from a population of patients that is similar to those being tested. The clinical applications for exercise testing include diagnosis of patients with chest pain syndromes, determination of disease severity, and prognosis in patients with known coronary artery disease, evaluation of arrhythmias, screening of asymptomatic patients, and evaluation of medical, surgical, and angioplastic therapy for coronary disease. In spite of studies involving thousands of patients, controversy exists regarding the diagnostic power of exercise testing. The large differences in reported accuracies are largely due to methodologic problems that have been encountered by various investigators. Clinicians should be made aware of these problems when reading the literature on ECG and radionuclide exercise testing. Such awareness will help them understand the limitations of these noninvasive procedures.(ABSTRACT TRUNCATED AT 400 WORDS)

Identification of anatomically extensive coronary artery disease by the exercise ECG ST segment/heart rate slope

To assess the ability of the ST segment/heart rate (ST/HR) slope to identify three-vessel coronary disease and the relationship between the ST/HR slope and the anatomic extent of disease as determined by the Gensini and Duke jeopardy scores, the exercise ECGs of 128 patients with stable angina were compared with findings at coronary cineangiography. A ST/HR slope greater than or equal to 6 microV/beat/min identified three-vessel coronary disease with a sensitivity of 93% compared with sensitivities of only 50% for early positive standard test responses (p less than 0.001) and 66% for markedly positive standard test responses (p less than 0.01). The negative predictive value of this ST/HR slope partition for three-vessel disease was 94%. Patients with ST/HR slopes greater than or equal to 6 who did not have three-vessel disease had anatomically more extensive obstruction than did patients with lower test values (mean Gensini score 43 +/- 5 vs 22 +/- 3, p less than 0.002 and mean jeopardy score 4.8 +/- 0.4 vs 3.0 +/- 0.3, p less than 0.01). Test performance of the calculated ST/HR slope exceeded that of a simplified index derived by dividing the total change in ST segment depression by the total change in heart rate. These findings demonstrate that a ST/HR slope greater than or equal to 6 is highly sensitive for the identification of three-vessel coronary disease and also identifies patients with anatomically severe obstruction. A ST/HR slope less than 6 makes three-vessel coronary disease or otherwise anatomically extensive coronary obstruction unlikely.

Lack of diurnal variation in maximal symptom-limited exercise test response in chronic stable angina

Exercise testing is widely used to evaluate the effects of anti-ischemic drugs. Many studies have reported good reproducibility when it is performed in the morning, but little information is available regarding the diurnal variation of exercise test response in patients with chronic stable angina. With the advent of new long-acting anti-ischemic drugs, it has become necessary to perform the exercise testing at various times of the day to determine the duration of action of a given drug. To examine the diurnal variation, exercise tests were performed on 41 patients, aged 53 to 75 years, with established chronic stable angina on 2 occasions 5 days apart at 10 A.M. and 4 P.M. on each day. On day 1, the mean +/- standard error of the mean exercise time was 5.0 +/- 0.4 minutes at 10 A.M. and 5.1 +/- 0.4 minutes at 4 P.M., and on day 5, it was 5.6 +/- 0.4 minutes at 10 A.M. and 5.5 +/- 0.4 minutes at 4 P.M. These values did not differ in statistical significance. Similarly, the time to the development of 1 mm of ST-segment depression did not show any statistically significant change during either test period on either day nor did maximal ST-segment depression. Heart rate at rest was 79 +/- 3 beats/min at 10 A.M., 81 +/- 3 beats/min at 4 P.M. on day 1 and 78 +/- 2 beats/min at 10 A.M. and 80 +/- 3 beats/min at 4 P.M. on day 5 (difference not significant). Similarly, no significant changes were observed in maximal heart rate or rate-pressure product at peak exercise.(ABSTRACT TRUNCATED AT 250 WORDS)

Influence of noise on wave boundary recognition by ECG measurement programs. Recommendations for preprocessing

In the international cooperative project entitled "Common Standards for Quantitative Electrocardiography" (CSE) systematic noise tests have been performed in order to compare measurement results of electrocardiographic computer programs under degraded operational conditions and to develop recommendations for preprocessing and measurement strategies. The influence of seven different high- and low-frequency noise types on the recognition of P, QRS, and T wave onsets and offsets was investigated. The analysis was performed on 160 electrocardiograms derived from two sets of 10 cases each, by eight electrocardiographic and six vectorcardiographic computer programs. The stability and precision of these programs were tested with respect to the results obtained (1) in the noise-free recordings and (2) by a group of five cardiologists who have analyzed the recordings previously in a Delphi reviewing process. Increasing levels of high-frequency noise shifted the onsets and offsets of most programs outward. Programs analyzing an averaged beat showed significantly less variability than programs which measure every complex or a selected beat. On the basis of the findings of the present study, a measurement strategy based on selective averaging is recommended for diagnostic ECG computer programs. However, averaging should be performed only if proper alignment and precise waveform comparison have been performed beforehand in order to exclude dissimilar complexes.

Description and evaluation of a method for computer analysis of the exercise electrocardiogram

The new approach to computer processing of exercise electrocardiography has been made easier by the development of microcomputers. Studies are necessary to validate analyzed electrocardiographic data for the diagnosis of ischemia. We describe and assess in this paper a new program for the analysis "on line" of 12 leads during effort. The program detects "normal QRS" and ectopic beats. Amplitude of R wave, length of QRS, ST level after a programmable delay from J point, ST maximal slope and amplitude of T wave are calculated and recorded every 15 sec in the 12 leads. In 200 exercise stress tests quantitative data provided by the processor were compared with visual analysis and with clinical data. ST level less than or equal to -0.8 mm and ST slope less than or equal to 1.2 mV/sec or ST level greater than or equal to +2.0 mm and ST slope less than or equal to 0.6 mV/sec were the best analyzed criteria for ischemia. Using these criteria, sensitivity increased from 86.6% by visual reading to 92% by computer analysis, without change in specificity (94%).

Does placebo improve indexes of effort-induced myocardial ischemia? An objective study in 150 patients with chronic stable angina pectoris

The effects of placebo were studied in 150 patients (135 men, 15 women) aged 42 to 75 years with stable exertional angina pectoris, using multistage graded exercise testing. Treadmill exercise, using on-line computer analysis of the electrocardiogram, was performed after a basal period, during which time the patients had no treatment for 2 weeks, and after 2 weeks of placebo therapy. Mean exercise time during no treatment was 6.0 +/- 0.2 minutes and during placebo was 6.1 +/- 0.2 minutes (difference not significant). Similarly, time to development of 1 mm of ST-segment depression of 4.0 +/- 0.2 minutes without treatment was 4.1 +/- 0.2 minutes after 2 weeks of placebo therapy (difference not significant). Placebo failed to show any effect on rest or maximal heart rate or on maximal ST-segment depression. It also failed to increase exercise tolerance or to improve other objective indexes of effort-induced myocardial ischemia in both single-and double-blind protocols in patients with stable exertional angina pectoris. Therefore, placebo control of antianginal drug trials that use exercise testing for evaluation of effect is unnecessary and can be omitted.

Maximum Shannon information content of diagnostic medical testing. Including application to multiple non-independent tests

The increase in Shannon information available from a diagnostic test associated with grading of the test results into many outcomes, rather than simply positive or negative, was examined to determine its upper limit as the number of test outcomes is increased indefinitely. Numerical methods were employed to find the optimal locations of outcome boundaries when a single normally distributed test variable is classified into 2, 3, 4, 5, 6, 8, 14, or 20 outcome categories. In each case Shannon information was computed for values of prior probability between 0.01 and 0.99 and for distances between the means in diseased and nondiseased populations ranging from 0.5 to 5.0 standard deviations. There is an important improvement in Shannon information as the number of outcomes defined is increased, but the increment in information diminishes rapidly with each additional category. A 20%-30% increment in information may be achieved with three outcomes instead of two. A further important increase in information occurs with four to seven outcomes, but beyond this the increment in inforation is negligible. The findings were similar over a wide range of prior probabilities and distances between the means. The analysis was extended to the case of multiple nonindependent tests by demonstrating their application to a Fisher discriminant function incorporating such tests. It was concluded that for normally distributed test variables: grading of test results significantly improves the information content of both single and multiple tests; the value of information content for 8-20 outcomes represents very nearly the maximum information content of a test; there is little value in using more than five to seven test outcomes; multiple grading should not be neglected for discriminant functions.

Specificity, sensitivity, and prevalence in the design of randomized trials: a univariate analysis

The number of cases needed to conduct a randomized trial is related to the sensitivity and specificity of a measurement indicative of a condition, to the prevalence of the condition, to the expected benefit of therapy (or other basis for change), and to the statistical precision desired. Sample size calculations frequently ignore sensitivity and specificity (at least qualitatively) probably because no simple formula is provided in the literature. Such a formula is included here. As an example, the number of patients required for a randomized clinical trial was calculated for a clinical outcome (nonfatal myocardial infarction or coronary artery disease death) used to detect atherosclerotic heart disease and is compared to the sample sizes required for each of three noninvasive diagnostic studies (exercise ECG ST depression, exercise LVEF reduction, and thallium myocardial imaging) performed for the detection of atherosclerotic heart disease. We calculated that the sample size should be much smaller when these diagnostic studies are employed compared to the clinical outcome, thereby offering the potential for reduced cost and complexity of a randomized clinical trial.

Improved accuracy of the exercise electrocardiogram: identification of three-vessel coronary disease in stable angina pectoris by analysis of peak rate-related changes in ST segments

Exercise electrocardiography has relatively poor specificity and predictive accuracy for 3-vessel coronary artery disease (CAD) when conventional diagnostic criteria are used. However, electrocardiographic evaluation using linear regression analysis of the heart-rate (HR)-related change in ST-segment depression (ST/HR slope) is reported to accurately distinguish patients with from those without CAD, and to accurately separate patients with 1-, 2- and 3-vessel CAD. To assess the applicability of this method and to compare it with conventional interpretation, retrospective evaluation of 50 patients in whom exercise electrocardiography and coronary cineangiography had been performed for suspected CAD was conducted using a modified ST/HR slope analysis limited to leads V5, V6 and aVF. Eighteen patients had 3-vessel, 22 had 2-vessel, 6 had 1-vessel and 4 had no CAD. Standard electrocardiographic criteria (1 mm or more of horizontal or downsloping ST depression) identified 3-vessel CAD with a sensitivity of 78%, specificity of 56% and positive predictive value of only 50%. Peak ST/HR slope criteria (greater than or equal to 6.0 microV/beat/min) identified 3-vessel CAD with a sensitivity of 78%, specificity of 97% and positive predictive value of 93%. The overall test accuracy using measured peak ST/HR slope was 90%, compared with 64% for standard ST-depression criteria. In conclusion, analysis of the peak ST/HR slope can greatly improve the diagnostic accuracy of exercise electrocardiography, and further prospective study of this method is indicated.

Reproducibility of multistage graded exercise testing in patients with chronic stable angina

Exercise testing is widely used for the diagnosis of ischaemic heart disease and for the evaluation of antianginal drugs. To assess reproducibility, analysis was carried out on 128 paired graded exercise tests from 103 patients performed at the same time of day and under identical conditions. Six different parameters were evaluated and compared between the basal test (no treatment) and the placebo test. During the basal period the mean (+/- SEM) exercise time to the development of angina was 6.0 (+/- 0.2) min and the 1 mm ST depression time was 4.1 (+/- 0.2) min. After 2 weeks of placebo the exercise time was 6.1 (+/- 0.2) min (P = NS) and the 1 mm ST depression time was 4.2 (+/- 0.2) min (P = NS). There was no significant difference between the resting or maximum heart rate on either test and the maximum ST segment depression (leads CM5 and CC5) was unaltered. In a second group of 17 patients where the basal tests were performed in the afternoon and the placebo tests in the morning, heart rate and ST segment were found to be reproducible but there was a significant difference in exercise time: 5.7 (+/- 0.7) min for the basal test and 8.3 (+/- 0.5) min for the placebo test (P less than 0.001); and of the 1 mm ST depression time: 2.7 (+/- 0.4) min for the basal test, and 5.4 (+/- 0.5) min for the placebo test (P less than 0.001). We conclude that exercise tests done under standardised conditions in the morning are highly reproducible in patients with chronic stable angina and therefore provide a valuable test for the evaluation of antianginal drugs.

Computer aided exercise electrocardiographic testing and coronary arteriography in patients with angina pectoris and with myocardial infarction

A set of electrocardiographic criteria for the diagnosis of coronary artery disease was evaluated in two different groups of patients examined by computer aided 12 lead exercise electrocardiographic stress testing and coronary arteriography. One group consisted of patients with severe angina pectoris and the other of patients who had suffered a myocardial infarction three years before the study. Angiographically determined categories of patients could be identified with satisfactory precision by the electrocardiographic criteria under test in the patients with angina pectoris but not in those with infarction. A new method of classifying patients on the basis of data from coronary arteriography improved the correlation with ST segment analysis compared with conventional classification.

Assessment of aortocoronary bypass grafting using exercise ST segment/heart rate relation

The maximal rate of progression of ST segment depression relative to increases in heart rate (maximal ST/HR slope) has recently been shown to be an accurate index of the presence and the severity of coronary heart disease in patients with angina. The value of this new exercise test was assessed in patients undergoing aortocoronary bypass. The maximal ST/HR slope and the results of coronary angiography were obtained in each of 46 patients before aortocoronary bypass surgery and in 26 of the 46 patients six months after the operation. At each stage of the investigation the maximal ST/HR slope detected without false results the absence and the number of significantly diseased vessels as shown by angiocardiography. As in previous findings the ranges of the maximal ST/HR slope showed no overlap between the four groups of patients: those with no significant disease and those with single, double, or triple vessel disease. In each of the 46 patients in whom the maximal ST/HR slope was determined before operation and three months afterwards the slope was lower after operation than before, indicating improvement. Follow up examinations showed that the maximal ST/HR slopes accurately detected the number of patent grafts used to bypass significantly diseased coronary arteries. Furthermore, the development of a significant narrowing or occlusion in any vein graft caused an increase in the maximal ST/HR slope which was equivalent to the value of single vessel disease. It is suggested that the maximal ST/HR slope may be used reliably in individual patients to indicate restoration of adequate blood supply to the myocardium after successfully aortorcoronary bypass surgery and the to detect in the period of six months after the operation the degree of severity of coronary heart disease whether it is caused by occlusion of the graft of significant disease of the coronary arteries.

Comparison of the multivariate analysis and CADENZA systems for determination of the probability of coronary artery disease

The accuracy of 2 discriminate systems for diagnosis of coronary artery disease (CAD), multivariate analysis (MVA) and Bayesian analysis (CADENZA), was evaluated in 113 patients undergoing electrocardiographic stress testing and coronary angiography. MVA uses weighting factors (F values) generated from our patient data, whereas CADENZA uses probabilities gleaned from an extensive review of the American literature. Overall accuracy was similar. MVA had a higher sensitivity for 1-vessel CAD (75 versus 33%), but CADENZA was better for determining the severity of CAD. The 2 systems provided posterior probabilities for disease that were highly correlated (r = 0.56; p less than 0.001). Both systems suggest the need for further testing based on the probability generated; herein lies their major strength. The application of such systems should help the clinician reach a diagnosis or make a decision as to management in a cost-effective manner.

Computer-derived treadmill exercise score quantifies the degree of revascularization and improved exercise performance after coronary artery bypass surgery

A computer-derived treadmill exercise score (TES) that quantifies the severity of the ischemic ST response to exercise was used to detect noninvasively graft occlusion or the progression of new lesions after coronary artery bypass surgery. Three months after surgery TES completely normalized in 68% of patients and improved by more than 70% in another 18% of patients, thus reflecting excellent improvement in exercise-induced ischemia in 87% of patients. Surgical results correlated well with completeness of revascularization as shown by repeat coronary angiography. When TES, done serially up to 4 years after surgery, remained unchanged, grafts were patent and no new critical lesions had occurred. Deterioration in TES always predicted either late graft occlusion or appearance of new, high-grade lesions in the native vessels. Thus TES provides a new, accurate method that quantifies the ischemic response to exercise and detects graft occlusion or new obstructive lesions in the native coronary arteries.

Exercise testing and training: clinical applications

The application of exercise in clinical cardiology continues to progress because of research findings. Advances have occurred in the applications, methodology and interpretation of exercise testing. Exercise training has been documented to have a place in the primary prevention of coronary heart disease. In regard to cardiac rehabilitation, both early ambulation and early discharge are safe and beneficial in patients with uncomplicated infarction, and a subsequent exercise program is at least as effective as other interventions. High intensity exercise training in the patient with heart disease may be necessary to cause changes in myocardial perfusion and performance, but it carries an increased risk.

Toward the optimal lead system and optimal criteria for exercise electrocardiography

To define the optimal lead system for exercise electrocardiography, data of the whole body surface potential distribution were analyzed in 25 normal subjects and in 25 patients with coronary artery disease at rest and during exercise. All patients had a normal electrocardiogram at rest. The sensitivity of the standard chest leads was 60 percent; it improved to 84 percent with the body surface map whereas both methods had a 100 percent specificity. On the basis of these data, and reports from other centers, it is concluded that a single bipolar lead from the right subclavian area to lead V5 is adequate in those laboratories that are restricted to testing subjects with a normal electrocardiogram at rest. In patients with a previous infarction or other abnormalities in the electrocardiogram at rest three (pseudo) orthogonal leads or several standard leads are necessary. Recommendations for optimal measurements from the exercise electrocardiogram are based on quantitative computer analysis of the selected leads in larger groups of patients. Best results were obtained with a combination of S-T amplitude, S-T slope and heart rate. The improvement in sensitivity from 50 percent with visual analysis to 85 percent with computer was similar to that obtained with body surface mapping. Changes of the P wave and QRS complex during exercise appeared to be of little diagnostic value. The pathophysiologic mechanisms that contribute to the changes of the electrocardiogram during exercise are discussed.