High-energy external defibrillation and transcutaneous pacing during MRI: feasibility and safety

Figures

• 

  Fig. 1. Defibrillator Pad Placement. Defibrillator pads were placed in the left-posterior axillary region (1) and right subclavian region (2). Temperature under the defibrillator pad was recorded continuously using fiber-optic temperature probes (arrows) with 0.01 °C resolution at approximately 10 samples/sec/channel

• 

  Fig. 2. Experimental Setup. The setup included: 1) an MRI-compatible, wireless 12-lead ECG monitor with defibrillation-overload protection and carbon ECG leadwires; 2) a variable isolation transformer, which was used to induce VF; 3) a defibrillator with radiolucent defibrillation pads; 4) a high-fidelity, fiber-optic temperature-monitoring system; and 5) an MRI compatible pulse-oximetry monitor with the sensor placed on the animal’s tail

• 

  Fig. 3. Functional Diagram of the MRI-compatible, Wireless, 12-lead ECG Module with Defibrillation-Overload Protection. The module is located on a patient table and provides both wireless and wired data links. It can move with the table between the MRI and fluoroscopy rooms. Data are transmitted via a bidirectional wireless link and router to a real-time review and display station, which is located in the control room. The module also provides optional gradient waveform and wired digital-data link, which were not used in this study. Inset: block diagram of the 12-lead ECG module. The filtering and conditioning module performs bandpass filtering using either diagnostic (0.05–150 Hz) or monitoring (0.5–50 Hz) frequency range. In addition, the module detects EMI in the input signals (e.g., ECG), using time-domain features of EMI patterns (e.g., derivative and amplitude) and discards the corresponding time sample or inserts an “EMI event” mark to guide the signal-processing algorithm [26, 27]

• 

  Fig. 4. ECG before and after Defibrillation during Active Scanning. ECG in lead V1 before defibrillation (a) and after defibrillation (b-d) during MRI scanning (balanced steady-state free precession sequence; bSSFP) in one of the animals (pig #6). Inset (e) shows detailed waveform morphology. Defibrillation was performed at time = 0

• 

  Fig. 5. Cardiac Pacing with Ventricular Capture. Sample ECG tracing during transcutaneous, constant-rate (non-demand) pacing at 60 bpm with ventricular capture in the magnet bore in one of the animals (pig #4) using monophasic, 20-ms, 150 mA pulses with rise/fall times ≤1 msec. The pacing was applied in a subset of two animals and produced reliable ventricular capture in both animals

• 

  Fig. 6. Mean and Max Temperature Changes after Defibrillation outside the Bore, inside the Bore and during Active Scanning in the Studied Animal Group (n = 6). The short-term temperature changes after defibrillation were limited to 1.5 °C in five of the six pigs. In the sixth animal, in which defibrillation was performed using pediatric pads, the max temperature increased significantly (> 2 °C) under one of the temperature probes during in-bore defibrillation. Moving the coil (covered in plastic) away from the pad reduced the post-defibrillation changes to approximately 1 °C

• 

  Fig. 7. Summary of Post-defibrillation Temperature Changes in the Studied Group. Changes in mean and max temperature (ΔT) 10 s after defibrillation outside the bore (control), inside the bore (in bore), and during bSSFP scan (scan). The temperature changes after defibrillation were slightly larger inside the bore and during scanning compared with outside the bore. Stars indicate significant differences compared with control measurements (p < 0.05, Wilcoxon matched pairs test)

• 

  Fig. 8. Temperature Changes under Defibrillation Pads during Continuous 2D bSSFP scan. The temperature registered by individual probes during a 6-min scan are shown with baseline temperature subtracted to calculate ΔT°C (a); thick lines indicate the maximum (Max T), mean (Mean T), and minimum (Min T) temperature changes. To determine the trajectory of the temperature changes, the lumped-capacitance single-term exponential heat-transfer model was applied in the following form: T = a − be−t/c, where T denotes temperature registered by the sensor and t is measurement time [24, 25]. The model satisfied acceptance criteria (RMSE < 0.03 and R2 > 0.98), and the model coefficients (a = 2.643; b = 2.295; c = 5.659) were used to predict the Max, Mean, and Min T trajectories beyond the 6-min interval (b). Note that the predicted Mean T trajectory is consistent (error: 0.1 °C) with the temperature registered under defibrillator pads after 11 min of similar bSSFP scanning in swine reported by Schmidt et al. (cross) [11]

• 

  Fig. 9. Examples of Skin Changes under the Adult and Pediatric Defibrillator Pads after Defibrillation. a After the removal of adult pads, the skin showed minimal or no visible changes. b After the removal of pediatric pads, the skin showed either no visible changes or, in one animal, redness and deeper lesions, particularly in the border zone. The affected skin area is demarcated by red dotted line

• 

  Fig. 10. MR Image Quality with the Defibrillator Turned On and Off. bSSFP images were obtained when defibrillator was switched off (a) and switched on (b). Note the reduction in image quality (due to increased noise level) when the defibrillator was switched on (b) compared with when it was switched off (a). The images were used to calculate SNR as previously described [11]. There was an approximate 10% reduction in SNR when the defibrillator was switched on. Bottom: image artifact caused by defibrillator pads (d, arrow) compared with the same image without the pads (c)