The derivation of the spatial QRS-T angle and the spatial ventricular gradient using the Mason-Likar 12-lead electrocardiogram

Overview of featurization techniques used in traditional versus emerging deep learning-based algorithms for automated interpretation of the 12-lead ECG

Automated interpretation of the 12-lead ECG has remained an underpinning interest in decades of research that has seen a diversity of computing applications in cardiology. The application of computers in cardiology began in the 1960s with early research focusing on the conversion of analogue ECG signals (voltages) to digital samples. Alongside this, software techniques that automated the extraction of wave measurements and provided basic diagnostic statements, began to emerge. In the years since then there have been many significant milestones which include the widespread commercialisation of 12-lead ECG interpretation software, associated clinical utility and the development of the related regulatory frameworks to promote standardised development. In the past few years, the research community has seen a significant rejuvenation in the development of ECG interpretation programs. This is evident in the research literature where a large number of studies have emerged tackling a variety of automated ECG interpretation problems. This is largely due to two factors. Specifically, the technical advances, both software and hardware, that have facilitated the broad adoption of modern artificial intelligence (AI) techniques, and, the increasing availability of large datasets that support modern AI approaches. In this article we provide a very high-level overview of the operation of and approach to the development of early 12-lead ECG interpretation programs and we contrast this to the approaches that are now seen in emerging AI approaches. Our overview is mainly focused on highlighting differences in how input data are handled prior to generation of the diagnostic statement.

Sex and race differences in J-Tend, J-Tpeak, and Tpeak-Tend intervals

To facilitate the precision of clinical electrocardiographic studies of J-to-Tpeak (JTp) and Tpeak-to-Tend (Tpe) intervals, the study investigated their differences between healthy females and males, and between subjects of African and Caucasian origin. In 523 healthy subjects (254 females; 236 subjects of African origin), repeated Holter recordings were used to measure QT, JT, JTp, and Tpe intervals preceded by both stable and variable heart rates. Subject-specific curvilinear regression models were used to obtain individual QTc, JTc, JTpc and Tpec intervals. Rate hysteresis, i.e., the speed with which the intervals adapted after heart rate changes, was also investigated. In all sex-race groups, Tpe intervals were not systematically heart rate dependent. Similar to QTc intervals, women had JTc, and JTpc intervals longer than males (difference 20–30 ms, p < 0.001). However, women had Tpec intervals (and rate uncorrected Tpe intervals) shorter by approximately 10 ms compared to males (p < 0.001). Subjects of African origin had significantly shorter QTc intervals than Caucasians (p < 0.001). Gradually diminishing race-difference was found for JTc, JTpc and Tpec intervals. JTc and JTpc were moderately increasing with age but Tpe/Tpec were not. Rate hysteresis of JTp was approximately 10% longer compared to that of JT (p < 0.001). In future clinical studies, Tpe interval should not be systematically corrected for heart rate and similar to the QT interval, the differences in JT, JTp and Tpe intervals should be corrected for sex. The differences in QT and JT, and JTp intervals should also be corrected for race.

Heart Rate Correction of the J-to-Tpeak Interval

Drug-induced changes of the J to T peak (JTp) and J to the median of area under the T wave (JT50) were reported to differentiate QT prolonging drugs that are predominant blockers of the delayed potassium rectifier current from those with multiple ion channel effects. Studies of drug-induced JTp/JT50 interval changes might therefore facilitate cardiac safety evaluation of new pharmaceuticals. It is not known whether formulas for QT heart rate correction are applicable to JTp and JT50 intervals. QT/RR, JTp/RR, and JT50/RR profiles were studied in 523 healthy subjects aged 33.5 ± 8.4 years (254 females). In individual subjects, 1,256 ± 220 electrocardiographic measurements of QT, JTp, and JT50 intervals were available including a 5-minute history of RR intervals preceding each measurement. Curvilinear, linear and log-linear regression models were used to characterize individual QT/RR, JTp/RR, and JT50/RR profiles both without and with correction for heart rate hysteresis. JTp/RR and JT50/RR hysteresis correction needs to be included but the generic universal correction for QT/RR hysteresis is also applicable to JTp/RR and JT50/RR profiles. Once this is incorporated, median regression coefficients of the investigated population suggest linear correction formulas JTpc = JTp + 0.150(1-RR) and JT50c = JT50 + 0.117(1-RR) where RR intervals of the underlying heart rate are hysteresis-corrected, and all measurements expressed in seconds. The established correction formulas can be proposed for future clinical pharmacology studies that show drug-induced heart rate changes of up to approximately 10 beats per minute.

Detection of T Wave Peak for Serial Comparisons of JTp Interval

Electrocardiogram (ECG) studies of drug-induced prolongation of the interval between the J point and the peak of the T wave (JTp interval) distinguished QT prolonging drugs that predominantly block the delayed potassium rectifier current from those affecting multiple cardiac repolarisation ion channel currents. Since the peak of the T wave depends on ECG lead, a “global” T peak requires to combine ECG leads into one-dimensional signal in which the T wave peak can be measured. This study aimed at finding the optimum one-dimensional representation of 12-lead ECGs for the most stable JTp measurements. Seven different one-dimensional representations were investigated including the vector magnitude of the orthogonal XYZ transformation, root mean square of all 12 ECG leads, and the vector magnitude of the 3 dominant orthogonal leads derived by singular value decomposition. All representations were applied to the median waveforms of 660,657 separate 10-s 12-lead ECGs taken from repeated day-time Holter recordings in 523 healthy subjects aged 33.5 ± 8.4 years (254 women). The JTp measurements were compared with the QT intervals and with the intervals between the J point and the median point of the area under the T wave one-dimensional representation (JT50 intervals) by means of calculating the residuals of the subject-specific curvilinear regression models relating the measured interval to the hysteresis-corrected RR interval of the underlying heart rate. The residuals of the regression models (equal to the intra-subject standard deviations of individually heart rate corrected intervals) expressed intra-subject stability of interval measurements. For both the JTp intervals and the JT50 intervals, the curvilinear regression residuals of measurements derived from the orthogonal XYZ representation were marginally but statistically significantly lower compared to the other representations. Using the XYZ representation, the residuals of the QT/RR, JTp/RR and JT50/RR regressions were 5.6 ± 1.1 ms, 7.2 ± 2.2 ms, and 4.9 ± 1.2 ms, respectively (all statistically significantly different; p < 0.0001). The study concludes that the orthogonal XYZ ECG representation might be proposed for future investigations of JTp and JT50 intervals. If the ability of classifying QT prolonging drugs is further confirmed for the JT50 interval, it might be appropriate to replace the JTp interval since with JT50 it appears more stable.

Common Genetic Variant Risk Score Is Associated With Drug-Induced QT Prolongation and Torsade de Pointes Risk: A Pilot Study

BACKGROUND

Drug-induced QT interval prolongation, a risk factor for life-threatening ventricular arrhythmias, is a potential side effect of many marketed and withdrawn medications. The contribution of common genetic variants previously associated with baseline QT interval to drug-induced QT prolongation and arrhythmias is not known.

METHODS

We tested the hypothesis that a weighted combination of common genetic variants contributing to QT interval at baseline, identified through genome-wide association studies, can predict individual response to multiple QT-prolonging drugs. Genetic analysis of 22 subjects was performed in a secondary analysis of a randomized, double-blind, placebo-controlled, crossover trial of 3 QT-prolonging drugs with 15 time-matched QT and plasma drug concentration measurements. Subjects received single doses of dofetilide, quinidine, ranolazine, and placebo. The outcome was the correlation between a genetic QT score comprising 61 common genetic variants and the slope of an individual subject's drug-induced increase in heart rate-corrected QT (QTc) versus drug concentration.

RESULTS

The genetic QT score was correlated with drug-induced QTc prolongation. Among white subjects, genetic QT score explained 30% of the variability in response to dofetilide (r=0.55; 95% confidence interval, 0.09-0.81; P=0.02), 23% in response to quinidine (r=0.48; 95% confidence interval, -0.03 to 0.79; P=0.06), and 27% in response to ranolazine (r=0.52; 95% confidence interval, 0.05-0.80; P=0.03). Furthermore, the genetic QT score was a significant predictor of drug-induced torsade de pointes in an independent sample of 216 cases compared with 771 controls (r²=12%, P=1×10^-7).

CONCLUSIONS

We demonstrate that a genetic QT score comprising 61 common genetic variants explains a significant proportion of the variability in drug-induced QT prolongation and is a significant predictor of drug-induced torsade de pointes. These findings highlight an opportunity for recent genetic discoveries to improve individualized risk-benefit assessment for pharmacological therapies. Replication of these findings in larger samples is needed to more precisely estimate variance explained and to establish the individual variants that drive these effects.

CLINICAL TRIAL REGISTRATION

URL: http://www.clinicaltrials.gov. Unique identifier: NCT01873950.

Computing the spatial QRS-T angle using reduced electrocardiographic lead sets

The 'spatial QRS-T angle' (SA) is frequently determined using linear lead transformation matrices that require the entire 12-lead electrocardiogram (ECG). While this approach is adequate when using 12-lead ECG data that is recorded in the resting supine position, it is not optimal in monitoring applications. This is because maintaining a good quality recording of the complete 12-lead ECG in monitoring applications is difficult. In this research, we assessed the differences between the 'gold standard' SA as determined using the Frank VGG and the SA as determined using different reduced lead systems (RLSs). The random error component (span of the Bland-Altman 95% limits of agreement) of the differences between the 'gold standard' SA and the SA values based upon the different RLSs was quantified. This was performed for all 62 RLSs that can be constructed from Mason-Likar (ML) limb leads I, II and all possible precordial lead subsets that contain between one and five of the precordial leads V1 to V6. The RLS with the smallest lead set size that produced SA estimates of a quality similar to what is achieved using the ML 12-lead ECG was based upon ML limb leads I, II and precordial leads V1, V3 and V6. The random error component (mean [95% confidence interval]) associated with this RLS and the ML 12-lead ECG were found to be 40.74° [35.56°-49.29°] and 39.57° [33.78°-45.70°], respectively. Our findings suggest that a RLS that is based upon the ML limb leads I and II and the three best precordial leads can yield SA estimates of a quality similar to what is achieved when using the complete ML 12-lead ECG.