Modeling and safety of magnetic spinal cord stimulation

Project summary/Abstract Spinal cord stimulation (SCS) has shown promise as an adjuvant therapy to conventional task-specific training to improve sensorimotor recovery after spinal cord injury. SCS is typically performed by implanting an electrode array into the dorsal epidural space. These electrodes create very focused electric fields (E-fields) that mainly recruit sensory nerves in the dorsal roots and dorsal spinal column. A more recent, less invasive SCS technique uses surface electrodes to pass currents through the patient’s skin (transcutaneous stimulation). However, electrical currents cannot penetrate deeply through biological tissues, and thus are unlikely to directly activate motor nerves in the ventral spinal column. Conventional SCS techniques may therefore not be optimal for motor rehabilitation, which may partly explain the clinically observed variability of treatment outcomes. Magnetostimulation is an emerging alternative neuromodulation technique using external coils to create time- varying magnetic fields that induce stimulating E-fields via Faraday induction. Magnetically induced E-fields can reach deeper into the body than electrode E-fields because biological tissues are virtually transparent to magnetic fields. Furthermore, there is no direct contact between the coil and the skin, which reduces stimulation of skin pain receptors and improves patient tolerability. Despite these advantages, magnetic spinal cord stimulation (mSCS) has not been modeled in detail and has not specifically been applied to motor rehabilitation. A few studies have been performed using conventional transcranial magnetic stimulation (TMS) figure-8 coils, but those have been found to be inefficient at creating E-fields that reach the cord. As a result, there is little safety data on mSCS, including the possibility of cardiac stimulation, secondary stimulation of critical peripheral nerves such as the vagus, and patient tolerability at increased power levels (higher power is required than in conventional TMS to reach the spinal cord, which is deeper than the cortex). We propose a comprehensive program of bioelectromagnetic modeling, hardware optimization and experimental validation to assess the safety, tolerability, and potential of mSCS for direct spinal cord stimulation for motor rehabilitation. This proposal leverages our expertise in magnetostimulation system design and modeling from our previous work on peripheral nerve and cardiac stimulation. We extend our modeling pipeline to mSCS and use it to identify an optimal coil design that directly stimulates the spinal cord while minimizing peripheral nerve stimulation and completely avoiding stimulation of the heart and of critical nerves such as the vagus. We construct an mSCS device based on the optimal coil design and deploy it in a healthy volunteer study to assess patient tolerability for slowly increasing stimulus intensities. This study will pave the way for larger studies focused on the clinical evaluation of mSCS for motor rehabilitation.