Tracking 12-lead ECG Waveforms during MRI-guided Interventional Procedures and Functional Exams

Background: Because scanner-generated electromagnetic interference obscures ECG waveforms, the role of ECG in the MRI environment has been limited to heart-rate tracking and synchronizing image acquisition with cardiac rhythm. MR-guided interventions and functional tests in patients with cardiovascular diseases are associated with heightened risk of acute ischemia and life-threatening arrhythmias, which require accurate tracking of subtle changes in the ECG waveform, particularly in the repolarization (ST-T) segment. However, the dynamic interaction between electrically conductive blood and strong magnetic field (the magnetohydrodynamic effect, MHE) distorts and obscures ECG waveforms, primarily the ST-T segment. We hypothesized that time-based ECG indices are more stable and reliable in the presence of MHE than amplitude-based indices. This hypothesis was tested using the activation-recovery interval (ARI), which is sensitive to acute ischemia and proarrhythmic electrophysiological abnormalities, representing an average action potential duration of the cardiomyocytes.

Methods: Pilot experiments were conducted in 4 healthy male volunteers (age: 30-55y) during cardiovascular magnetic resonance (CMR) imaging using a 3T scanner. Twelve-lead ECG was acquired using a high-fidelity, MRI-compatible system (PinMed, Pittsburgh, PA); QRS complexes and ARI were identified by a validated ECG-analysis algorithm and verified by an experienced ECG analyst (Fig. 1). The data were analyzed in 1-min intervals, with the patient table in home position (baseline); at 30, 60, 90, and 120 cm table shift; and in the scanner’s isocenter during active CMR scanning. ARI was estimated in each lead and in the root-mean-square (RMS) curve constructed from all leads.

Results: QRS complexes and ARI were identifiable during the entire exam, including active scanning; significant ECG distortions due to MHE were observed at table positions ≥90 cm. ARI was identifiable during the entire exam and correlated with the changes in RR intervals (Fig. 2), as expected, with the greatest correlation in lead V3 (r = 0.66, p < 0.001, Fischer test).

Conclusion: In the pilot experiments, time-based ECG indices were more stable and reliably identifiable than amplitude-based indices in 12-lead surface ECG during active scanning in a 3T scanner. Further research is warranted to elucidate the potential of ARI tracking for monitoring changes in cardiac electrophysiology and for detecting acute ischemia during MR-guided interventional procedures and functional tests.

Fig. 1. Sample ARI measurements in the root-mean-square (RMS) ECG tracing in Subject #1 at baseline (A) and during active scan (B). Sampling interval: 2 ms.
Fig. 2. Dynamics of RR intervals (A) and ARI (B) in lead V3 in the studied group (median, 25-75% range) at baseline, 30 cm, 60 cm, 90 cm, 120 cm, and at the isocenter during scanning. The dynamic patterns of ARI and RR were correlated (p < 0.001).