Method for the evaluation of the performance of a multi-channel nerve cuff electrode coupled with a transcutaneous transfer array

Abstract

A multi-channel nerve cuff electrode can be used to elicit selective contractions of different muscles depending on the position of the stimulating electrode. The stimulation current evokes action potential in the nerve fibres which is passed on to the attached muscles and results in contraction. Depending on the configuration of the fascicles, the different channels of the nerve cuff electrode will preferably address different muscles, allowing for selective control of movement. A method for the evaluation of a system for the selective stimulation of the gastrocnemius and the tibialis anterior muscles was developed. The system consisted of a multi-channel nerve-cuff electrode wrapped around the sciatic nerve and a coupling array, which allowed for the resistive transfer of current through the skin. These components were connected and could be assessed separately through a headstage port. The system was implanted in 10 female Sprague Dawley rats and tested under in vivo conditions. Measurements were performed to determine the current distribution on the channels of the transfer array as well as the selectivity of the different channels in the nerve cuff electrode with respect to stimulating only the gastrocnemius muscle or only the tibialis anterior muscle. Additionally measurements of the entire system were performed. The experiments were performed over a duration of 12 weeks after surgery. Two models were developed, to estimate the necessary array performance using the results from the measurements of the cuff electrode. Most subjects did not survive for the duration of the experiment. The mean lifespan amounted to 33 days (SD = 25 days) after surgery, not taking into account the animals that died on the day of surgery. Only one animal survived for the intended duration. Due to the rats movement mechanical failure was observed in 6 out of 8 animals, reducing the amount of performed experiments further As the mechanical failure usually only prevented measurements of a single component, the remaining system was still examiined. The mean time for the first mechanical failure to be detected was 24 days (SD= 11 days). Crosstalk coefficients between 73% and 115% were observed, where values above 100% mean, that the secondary channel received more current than the primary channel (lower numbers signify a better performance). The measurements of the nerve cuff electrode were used to estimate the necessary crosstalk coefficients to achieve selective behaviour in the full system. The more optimistic model, which assumed two separate stimulation sites to be fully independent, estimated the necessary crosstalk coefficient to be below 52% for any selectivity to be possible and below 37% for the coupling array to no longer be the limiting component for the performance. The second model which assumed a linear interaction between two stimulation sites estimated crosstalks coefficients below 23% to be necessary for the function of the system. Stimulation with the full system was possible, however the measurements of the full system did not show any selective behaviour, which matches the expectations from the results of the separate components. The sample size was drastically reduced due to the high amount of mechanical failures and the short life expectancy of the test subjects. The implementation in an in vivo subject proved to be technically problematic due to the size and stability of the implant. The performance of the electrode array was worse than under previously published ex vivo conditions and was insufficient for the system as a whole to perform selectively. However we were able to show selective behaviour of the nerve cuff electrode and develop a method to estimate the necessary crosstalk coefficient of the input signal. The results of the cuff measurements are in line with the work of previous authors, who performed similar experiments using steering currents. This is a technique that should be strongly considered in the further development. Further in vivo experiments should be postponed until the performance of the electrode array matches the thresholds defined by the models using the data from the cuff measurements. In addition to improving the performance of the system, the biocompatibility of the implant needs to be improved, so that the test subjects can survive for the intended duration. This would require a redesign, ideally including a reduction in size as well as eliminating elements protruding the skin. The performance of the nerve cuff electrode can be studied more thoroughly in a smaller implant to provide more precise values for the required crosstalk coefficient benchmarks without the need of implanting the entire system.

Read the full publication for complete study details.

Source: Ogrodnik, E. (2022). Method for the evaluation of the performance of a multi-channel nerve cuff electrode coupled with a transcutaneous transfer array (Doctoral dissertation, Technische Universität Wien).