A Methodological Framework for the Efficient Characterization of Peripheral Nerve Stimulation Parameters

Objective. Peripheral nerve stimulation (PNS) for restoration of movement and somatosensation requires precise manipulation of neural activation. The intensity of neural response is primarily modulated via pulse amplitude (PA) and pulse width (PW), which have differential effects on axon recruitment that could be harnessed for improved selectivity. However, simultaneously modulating both parameters is rarely done because of the time required to map the two-dimensional space. In this paper, we propose and clinically validate an efficient method to characterize multiple intensities in the PA-PW stimulation space for both motor and perceptual sensory applications. We also elucidate the mechanisms driving distinct activation patterns within the two-dimensional space and justify the practical application of intensity-matched stimulation across it. Approach. We applied PNS through cuff electrodes implanted in one participant with a spinal cord injury and two participants with upper limb loss to generate equal muscle activation and equal perceptual intensity contours, respectively, across the full functional intensity range in the PA-PW parameter space. Strength-duration (SD) curves were mapped to the contours and assessed for goodness of fit with varying sample point selection. Using finite element modeling of the human nerve and activation simulations, stimulation across the PA-PW parameter space was evaluated for differences in recruited axon populations. Main results. From the clinical work, SD curves fit all levels of motor activation and perceptual sensory intensity highly accurately (median R ² = 0.996,0.984 respectively). SD curves of all intensities can be reliably and accurately estimated using only two points given the two points are sufficiently distanced. A novel method for efficiently characterizing the PA-PW space utilizing the SD curve is proposed, including a metric to determine if the sampled points are likely to produce an accurate estimate. From the in silico work, intensity-matched high PW and high PA stimulation were shown to activate unique subsets of axons, with high PA stimuli preferentially activating large diameter motor and sensory axons further away from the contact. Significance. This work provides important justification and guidance for utilizing SD curves to efficiently define a two-dimensional stimulation region for clinical functional motor and sensory PNS. Application of the characterization method proposed in this study could yield improved selectivity and resolution of PNS for decreased fatigue, improved fine motor control, and unique percept generation.