Abstracts

Rationale: The focus of this work was in the area of gradient induced currents effects on the vagus nerve stimulator in MRI. VNS Therapy is used in the treatment of epilepsy while MRI is commonly used in the diagnosis and treatment of epilepsy. The strong gradient fields present in an MRI scanner can produce induced currents near the electrodes of the vagus nerve stimulator that could interfere with intended pulse generator signals or be of magnitude that introduces safety concerns for patients. It is therefore necessary to understand/quantify such induced currents near the electrodes of vagus nerve stimulator. Methods: A 3T Siemens MR scanner was used for the experimental measurements using an Echo Planar Imaging (EPI) sequence. B-dot probe measurements were used to determine the location of maximum gradients fields within the bore. Measurements were made with the VNS device (n = 6) in worst case configurations within a saline phantom at the location of maximum gradient fields with the stimulation turned off . Measurements were also made with the VNS device in the on configuration to evaluate the effect of the gradients on the output. In addition, confirmatory numerical simulations using code based on the low frequency impedance method and virtual family models were performed. The VNS was modeled in representative clinical regions to determine the induced current levels near the electrodes for comparison to measurements of the device in the off case. The measured maximum gradient waveform was used to generate the incident magnetic field for the simulation code. The Fourier transform of the waveform revealed three major frequency components. Simulations were then performed at these three frequencies and linear superposition was used to find the maximum gradient-induced current level. Results: A strength duration curve was established for the lowest VNS output (0.25 mA) for effective durations 10 sec - 10 msec. Measurements of induced currents of all devices tested were found to be under this strength duration curve. Measurements demonstrated negligible effect on the VNS output when the stimulation remained on . Numerical simulations confirmed these low levels of induced currents for the off case. Figure 1 shows the normalized induced current densities and magnitudes for the required incident field. The top three diagrams in Figure 1 show the current density levels inside the complete virtual family boy model at the three simulation frequencies, while the bottom three diagrams show the induced current magnitudes near the lead electrodes. Table 1 provides the peak induced current values found for the adult male virtual family model. For all cases the MRI gradient induced current levels near the lead electrodes were very low. Conclusions: Measurements in a 3T scanner using a saline-filled phantom with the VNS device in worst case configurations is supported by numerical modeling. Results of both indicate that maximum gradient induced currents are less than the lowest expected VNS output current.