Safety of 3 Tesla MRI in active peripheral nerve field stimulation device for facial pain

INTRODUCTION: Peripheral nerve field stimulation (PNFS) for facial pain delivers subcutaneous electrical stimulation to reduce pain. Functional MRI (fMRI) can be used to characterize central effects of neuromodulation techniques such as deep brain stimulation (DBS) and spinal cord stimulation (SCS). However, the safety and utility of MRI in patients with PNFS has not been established, limiting both clinical MRI use and the application of fMRI in this population. This study evaluated the MRI safety and feasibility of imaging an active SCS implant used for PNFS in patients with facial pain; and defined sequence parameters for concurrent blood-oxygenation-level-dependent (BOLD) fMRI acquisition. METHODS: An anthropomorphic 3D-printed phantom filled with tissue-mimicking gel and fitted with an SCS implant replicating a patient with PNFS was used for in vitro safety testing. Two phantom experiments evaluated the relationship between (i) head specific absorption rate (SAR), (ii) time-averaged positive radiofrequency magnetic field component (B1+rms), and maximal temperature rises at critical locations (i.e., distal lead electrodes, cranial coiling, and implantable pulse generator) across clinical and research-based structural and fMRI sequences. For validation, a PNFS patient was scanned using localizer, T1-weighted Magnetization-Prepared Rapid Gradient Echo (T1w MPRAGE), and BOLD fMRI sequences informed by phantom experiments. RESULTS: FMRI during active PNFS is safe under specific conditions, with temperature increases remaining below the 2°C threshold at all monitored locations. Heating had a stronger relationship with head SAR (higher adjusted coefficient of determination (𝑅2) value) than B1+rms, particularly at distal lead electrodes. These in vitro findings informed selection of safe fMRI protocols for in vivo scanning. A patient (n=1) underwent MRI with no device- or patient-related adverse events. Successful fMRI acquisition was achieved, demonstrating engagement of pain-related regions in the patient. CONCLUSION: Phantom testing confirmed the safety and feasibility of MRI with an active SCS device configured for facial PNFS. These findings, specific to the tested conditions, underscore the need for context-specific safety evaluations to enable safe MRI in such implantable medical devices.