The impact of electrode design on nerve stimulation
Vagus nerve stimulation has emerged as an effective treatment for a variety of disorders including epilepsy, stroke, and spinal cord injury (SCI). The stimulation parameters used, as well as the method of nerve activation, greatly affect therapeutic efficacy. Clinically, the typical stimulation method involves two helical electrodes wrapped around the cervical branch of the nerve. However, alternative designs may offer advantages such as ease of fabrication and implantation, and newly developed non-invasive techniques may eliminate surgery altogether. Despite these possibilities, there has been little investigation into how these design choices impact therapeutic efficacy. This dissertation seeks to address this issue by performing parametric characterizations and comparisons of multiple nerve stimulation techniques. First, a combination of computational modeling and in vivo testing was used to compare and contrast various cuff electrode designs. The results revealed how multiple geometric parameters of the implant affect nerve activation and demonstrated the viability of a new, flat electrode design. Next, stimulation via an implant was compared to two non-invasive methods: transcutaneous cervical VNS (tcVNS) and transcutaneous auricular VNS (taVNS). A physiological biomarker of vagus nerve stimulation, the Hering-Breuer (HB) reflex, was used to determine nerve activation thresholds for each type of stimulation. The results of this study confirmed that VNS delivered with an implanted cuff robustly engages the HB reflex. Additionally, while tcVNS is capable of activating the HB reflex, and thus the nerve, it is unlikely to do so at clinically relevant parameters. Lastly, no activation of the HB reflex was observed with taVNS, revealing a clear difference between the cervical and auricular branches of the vagus nerve. Besides VNS, several other nerve stimulation applications have emerged as effective treatments for multiple disorders. While it is possible the findings from the previous studies will translate to these other applications, key differences between the nerves or electrode designs may lead to differing results. Computational models, as the one developed in the first study, could be used to predict how these differences would affect therapeutic outcomes, but they require access to specialized software and creation of a new model for each application. To simplify this process, we developed a parameterized model using open-source software with an online front-end interface (https://nervestimlab.utdallas.edu). We demonstrated the accuracy of the model by comparing its results to in vivo data reported in previous literature. Its accuracy, simple interface, and large number of adjustable parameters indicate that it is a useful tool for various applications. Preclinically, VNS has proven to be an effective treatment for SCI. However, given the myriad of cardiovascular consequences of SCI and the direct effects of VNS on cardiac function, it is essential to assess the safety of VNS following SCI. We measured cardiac responses to VNS in rats with chronic SCI and obtained initial evidence that VNS as used to treat SCI appears safe. The experimental results in this dissertation, as well as the newly developed computational model can be used to make better informed decisions about how to apply nerve stimulation for a variety of applications.