Alternative Lead Systems for Diagnostic Electrocardiography: Validation and Clinical Applicability

Abstract: The standard 12-lead electrocardiogram (ECG) remains one of the most important and most frequently used tools for diagnosing cardiac diseases, although several different examination modalities in cardio¬logy have been developed over the years. The standard ECG uses 10 electrodes placed on well-defined positions on the body, 6 on the torso and 4 distally on the limbs. Both industry and academia have invested many years in development of the criteria used to interpret the “diagnostic” standard ECG, and the waveform patterns are taught in medical school. In several situations, however – such as during long-term ECG monitoring or stress testing – use of the electrode positions of the standard ECG is not optimal because of the abundance of noise. In these situations, the limb electrodes must be placed proximally, often even on the torso, and the Mason-Likar (M-L) positions are commonly used. Interference with other clinical procedures, such as echocardiography, can also constitute a problem. An ECG-recording system with fewer electrodes and without any electrodes on the limbs that provides a 12-lead ECG similar to the standard ECG would be valuable. The so-called EASI system uses only 4 recording electrodes in easily determined locations on the torso from which the full 12-lead ECG can be derived. The 12-lead ECG derived from the EASI system has been evaluated in adults in several clinical situations. Physicians who use ECGs in their day-to-day work are often not aware of the differences between 12-lead ECGs recorded from standard versus alternative electrode positions, and they might use criteria developed for the standard ECG when interpreting an ECG obtained from an alternative lead system. This can lead to misinterpretation with the risk of potentially serious consequences for the patient. Optimizing the proximal positions for better concordance with the standard ECG would be of great value for improved diagnostic performance. A version of the “Lund” (LU) lead system has been reported to agree better with the standard lead system than does the M-L lead system, with regard to both Q-wave width and QRS frontal plane axis. To develop a uniform convention for ECG recording, i.e. both for diagnostic ECG and for monitoring, a recording must produce waveforms that have morphologies approximating those obtained with standard ECG and that has noise immunity close to that of M-L. The overall objectives of this thesis were 1) to further validate the EASI system to gain more knowledge about the agreement between EASI-derived and standard 12-lead ECGs, and 2) to investigate the possibility of optimizing the positions of proximally placed limb electrodes. EASI studies In Study I, age-specific transformation coefficients were determined for use in deriving 12-lead ECGs from the EASI signals. The agreement of the waveforms between simultaneously recorded standard and EASI-derived 12-lead ECGs in children (healthy and with various cardiac diagnoses) was studied. For children, it was better to use age-specific transformation coefficients than adult coefficients. The agreement between standard and EASI-derived ECGs was mostly good. In Study II, the intrareader variation of interpretations of 2 standard 12-lead ECGs was compared with the variation of interpretations of standard versus EASI-derived 12-lead ECGs in children (Study I population). The variation of the interpretation of standard versus EASI-derived ECGs was only slightly larger than the intrareader variation of interpretations of standard ECGs. In Study III, the amplitudes of myoelectric noise and baseline wander were compared between simultaneously recorded EASI-derived and M-L 12-lead ECGs in healthy adults. Overall, the 2 lead systems had similar susceptibilities to baseline wander, but EASI was less susceptible than M-L to myoelectric noise. In Study IV, differences in the estimated size of myocardial infarction (MI), as assessed by Selvester scores, were compared between standard and EASI-derived 12-lead ECGs among patients who had had an episode of chest pain suggestive of an acute coronary syndrome. These scores were also compared with MI size measured by cardiac magnetic resonance imaging (MRI). Estimated MI size did not differ significantly between the 2 lead systems, but neither the correlation nor the agreement between MRI and either of the 2 lead systems was very strong. Study to optimize the proximal positions of the limb electrodes In Study V, waveforms from the LU and M-L systems were compared with those from standard ECGs with regard to the QRS axis in the frontal plane and QRS changes of inferior MI. The noise immunities of the standard, LU, and M-L systems were also compared. LU produced ECG waveforms that more closely resembled those obtained with standard ECG than did M-L. The LU system was more noise-immune than was the standard system, and the noise immunities of the LU and the M-L systems were comparable.