Peripheral nerve stimulation informed design of a high-performance asymmetric head gradient coil

Purpose: Peripheral nerve stimulation (PNS) limits the image encoding performance of both body gradient coils and the latest generation of head gradients. We analyze a variety of head gradient design aspects using a detailed PNS model to guide the design process of a new high-performance asymmetric head gradient to raise PNS thresholds and maximize the usable image-encoding performance.

Methods: A novel three-layer coil design underwent PNS optimization involving PNS predictions of a series of candidate designs. The PNS-informed design process sought to maximize the usable parameter space of a coil with <10% nonlinearity in a 22 cm region of linearity, a relatively large inner diameter (44 cm), maximum gradient amplitude of 200 mT/m, and a high slew rate of 900 T/m/s. PNS modeling allowed identification and iterative adjustment of coil features with beneficial impact on PNS such as the number of winding layers, shoulder accommodation strategy, and level of asymmetry. PNS predictions for the final design were compared to measured thresholds in a constructed prototype.

Results: The final head gradient achieved up to 2-fold higher PNS thresholds than the initial design without PNS optimization and compared to existing head gradients with similar design characteristics. The inclusion of a third intermediate winding layer provided the additional degrees of freedom necessary to improve PNS thresholds without significant sacrifices to the other design metrics.

Conclusion: Augmenting the design phase of a new high-performance head gradient coil by PNS modeling dramatically improved the usable image-encoding performance by raising PNS thresholds.