Towards Wearable Neuroprostheses for Hand Function Restoration

Abstract

Regaining hand function is the primary demand among people with spinal cord injury. Among the interventions available, functional electrical stimulation is preferred for its ease of use and non-invasiveness. However, realizing a complete hand function is still challenging, as these systems must overcome the physiological limitations of non-invasive stimulation, including selectivity and discomfort. Hence, this thesis aims to improve the applicability of wearable stimulation systems to facilitate complex hand function tasks by addressing these limitations. Also, to improve them through electrode designs and material technologies, which can offer improved stimulation performance favoring prolonged usage. Firstly, a scanning routine identified the motor points of flexors and extensors across the forearm. Later, a machine learning-based clustering algorithm grouped these motor points, which aided in deriving a generalized catalog that streamlined electrode placements. Secondly, optimal stimulation parameters that elicited controlled and comfortable levels of muscle contraction were identified using isokinetic dynamometry. Additionally, the influence of stimulation-induced fatigue was assessed under tetanic contraction. By varying the stimulation parameters and the corresponding motor points of the target muscles, the likelihood of personalized digit/wrist control was demonstrated Furthermore, the potential for electrode designs with a simple 2D geometry having improved stimulation performance was shown using both model-based and experimental assessments. Lastly, a conformable electrode array-based sleeve, suitable for the stimulation of forearm muscle groups was fabricated using a multi-layered screen-printing process. Herewith, the conductivity of a silicone-based elastomer was altered by the addition of carbonaceous material. Further characterization studies revealed that electrodes with an optimal ratio of carbon black infused within these elastomers tend to maintain their stimulation performance after several stretching cycles. By selectively eliciting digit control through motor point-based stimulation, exerting controlled digit forces by varying stimulation parameters, and synergistically activating muscle groups via distributed electrodes, the viability for realizing complex hand function tasks was demonstrated. Furthermore, this thesis also establishes the potential for improving stimulation performance by modifying the electrode geometry and the viability of fabricating a conformable electrode array-based sleeve. Ultimately, utilizing these electrode designs and advanced fabrication techniques can pave the way for comfortable and cost-effective wearable stimulation systems.