Background

The accuracy of Electrocardiogram (ECG) gating for synchronization of MR scanner image acquisition and cardiac electrical activity is of great importance for acquiring high-quality Cardiac Magnetic Resonance (CMR) images free of motion artefacts. The distortion of ECG traces by Magnetohydrodynamic Voltages (VMHD) induced by interaction between the MRI static magnetic field (B0) and rapid left-ventricular blood ejection during systole can lead to false and/or intermittent QRS complex detection and images with severe motion artefacts [1]. We hypothesized that an optimized electrode placement for the reduction of induced VMHD could be derived based on a thoracic model to increase the accuracy of QRS complex detection.

Methods

A vector model based on thoracic geometry [2] was calibrated using 12-lead ECGs recorded in four subjects in a GE 3T scanner to estimate VMHD distributions on the thorax. 4-lead ECG electrode placement was then optimized to: (1) minimize VMHD magnitude and (2) reduce displacement from the SA node for maximizing QRS complex amplitude (Figure 1a,b). A gradient-descent optimization routine was utilized to predict the optimal 4-lead ECG placement based on angular displacement and heart/aorta geometry (Figure 1c,d). Model results were then validated using five healthy subjects. Sensitivity (Se) and Positive Predictability (+P) rates for detection of R-waves were compared between conventional and MHD-suppressed lead placements for single-lead QRS complex detection [3].

Figure 1
figure 1

Development of electrode placement recommendations for increased accuracy in QRS complex detection and MRI gating.

Full size image

Results

A 43.41% reduction in VMHD during the S-T segment (Figure 1f) was observed in ECGs using the MHD-suppressed placement relative to the conventional placement, while preserving the QRS complex (Figure 1e), resulting in an average increase in the Se and +P rate of 14.22% and 15.48%, respectively (Figure 1e-g). Rpeak amplitude inside the MRI in the MHD-suppressed placement had <5% deviation from the standard placement outside of the MRI (Figure 1e). As compared to the conventional electrode placement (Figure 1c-d), MHD suppression may result from decreased visibility of the aorta through the lungs at the MHD-suppressed placement.

Conclusions

Electrode placement recommendations were computed and validated in a 3T MRI, illustrating an increased accuracy in QRS complex detection using the MHD-suppressed placement.