<p>Deep brain stimulation (DBS) is an established surgical therapy for movement disorders. The mechanisms of action of DBS remain unclear, and selection of stimulation parameters is a clinical challenge and can result in sub-optimal outcomes. Closed-loop DBS systems would use a feedback control signal for automatic adjustment of DBS parameters and improved therapeutic effectiveness. We hypothesized that evoked compound action potentials (ECAPs), generated by activated neurons in the vicinity of the stimulating electrode, would reveal the type and spatial extent of neural activation, as well as provide signatures of clinical effectiveness. The objective of this dissertation was to record and characterize the ECAP during DBS to determine its suitability as a feedback signal in closed-loop systems. The ECAP was investigated using computer simulation and <italic>in vivo</italic> experiments, including the first preclinical and clinical ECAP recordings made from the same DBS electrode implanted for stimulation. </p><p>First, we developed DBS-ECAP recording instrumentation to reduce the stimulus artifact and enable high fidelity measurements of the ECAP at short latency. <italic>In vitro</italic> and <italic>in vivo</italic> validation experiments demonstrated the capability of the instrumentation to suppress the stimulus artifact, increase amplifier gain, and reduce distortion of short latency ECAP signals.</p><p>Second, we characterized ECAPs measured during thalamic DBS across stimulation parameters in anesthetized cats, and determined the neural origin of the ECAP using pharmacological interventions and a computer-based biophysical model of a thalamic network. This model simulated the ECAP response generated by a population of thalamic neurons, calculated ECAPs similar to experimental recordings, and indicated the relative contribution from different types of neural elements to the composite ECAP. Signal energy of the ECAP increased with DBS amplitude or pulse width, reflecting an increased extent of activation. Shorter latency, primary ECAP phases were generated by direct excitation of neural elements, whereas longer latency, secondary phases were generated by post-synaptic activation.</p><p>Third, intraoperative studies were conducted in human subjects with thalamic DBS for tremor, and the ECAP and tremor responses were measured across stimulation parameters. ECAP recording was technically challenging due to the presence of a wide range of stimulus artifact magnitudes across subjects, and an electrical circuit equivalent model and finite element method model both suggested that glial encapsulation around the DBS electrode increased the artifact size. Nevertheless, high fidelity ECAPs were recorded from acutely and chronically implanted DBS electrodes, and the energy of ECAP phases was correlated with changes in tremor. </p><p>Fourth, we used a computational model to understand how electrode design parameters influenced neural recording. Reducing the diameter or length of recording contacts increased the magnitude of single-unit responses, led to greater spatial sensitivity, and changed the relative contribution from local cells or passing axons. The effect of diameter or contact length varied across phases of population ECAPs, but ECAP signal energy increased with greater contact spacing, due to changes in the spatial sensitivity of the contacts. In addition, the signal increased with glial encapsulation in the peri-electrode space, decreased with local edema, and was unaffected by the physical presence of the highly conductive recording contacts.</p><p>It is feasible to record ECAP signals during DBS, and the correlation between ECAP characteristics and tremor suggests that this signal could be used in closed-loop DBS. This was demonstrated by implementation in simulation of a closed-loop system, in which a proportional-integral-derivative (PID) controller automatically adjusted DBS parameters to obtain a target ECAP energy value, and modified parameters in response to disturbances. The ECAP also provided insight into neural activation during DBS, with the dominant contribution to clinical ECAPs derived from excited cerebellothalamic fibers, suggesting that activation of these fibers is critical for DBS therapy.</p> / Dissertation

Current cochlear implant systems use fast pulsatile stimulation to deliver the temporal modulations of speech and to, potentially, improve the neural representation of such modulations by restoring the independence of neural firing. The realization of these benefits may vary with other pulse rate-dependent temporal interactions that occur at the neural membrane, e.g., per(i)stimulatory adaptation and its post-stimulatory or forward masking effects. This study attempted to characterize adaptation and recovery of the electrically evoked compound action potential (ECAP) using probe pulses delivered within and following brief (100 ms) high-rate masker (1800 pps) pulse trains at various current levels in adults and children. With this stimulus paradigm, the ECAP amplitude typically achieved a steady state during the course of pulse train stimulation. The ECAP amplitude at steady state was, on average, a similar proportion (50-70%) of the amplitude at onset for various stimulus levels and in both age groups. However, long-term adaptation effects, evidenced by the decrease in onset ECAP amplitude, were greater in adults particularly at lower levels in the ECAP dynamic range. Instances of alternation in ECAP amplitude were seen at stimulus levels that were higher in the ECAP dynamic range. The forward masking effects of pulse train stimulation were quantified by the ECAP amplitude in response to a subsequent probe pulse normalized by the response to the same pulse presented alone. Pulse train forward masking increased with the level of the masker pulse train and decreased with the level of the probe stimulus. The recovery of the ECAP for probes that were lower in level than the masker pulse train was incomplete at 600 ms after masker offset, consistent with long-term cumulative effects observed in the response to the probe alone. Masker pulse trains that are lower in level than the probe pulse produced proportionally small decrements in the ECAP amplitude with complete recovery within 250 ms of pulse train offset particularly in adults. ECAP recovery of a probe preceded by a masker pulse train of equal level followed a monotonic or non-monotonic pattern consistent with a hypothesis of both adaptation and facilitation occurring with pulse train stimulation. The various patterns of recovery may attest to the occurrence of more than a single process in the same subset of nerve fibers or in different fibers. We hypothesize that the variations in the recovery patterns may be attributable to individual differences in the status of the auditory nerve and possibly, the variations in temporal interactions across the spatial domain at different stimulus levels. Finally, the probe-evoked ECAP amplitude at steady state in children and briefly, e.g., 20 ms, after pulse train offset in both age groups could be predicted by the ECAP amplitude in response to the same probe pulse when preceded at a brief interval (1.2 or 2 ms) by a single masker pulse of the same level as the masker pulse train. Further investigation may reveal if the observed differences in neural responsiveness to pulsatile stimulation, among individuals account for differences in psychophysical measures, including speech perception and whether there may be an "optimal" neural output that could be evoked by an individually "optimized" signal.

The purpose of this study was to investigate how design changes in Cochlear Nucleus cochlear implants (CIs) (CI24M, CI24R, CI24RE and CI422) affected electrode impedance and ECAP measures, and to determine if these design changes affected post-lingually deafened adults and children with congenital hearing loss in a similar way. Results of this study showed that electrode impedance was inversely related to the area of the electrode contacts in the array: lowest for the full-banded CI24M CI and highest for adults who used the CI422 device which has the smallest electrode contacts of all four devices. The noise floor of the NRT system likely plays a significant role in the finding that CI users with older devices (the CI24M, and CI24R CIs) had higher ECAP thresholds than individuals with the CI24RE electrode array. The position of the electrode array in the cochlea was also found to have a significant effect on ECAP measures. CI users with modiolar hugging (the CI24R and CI24RE CIs) electrode arrays were found to have lower ECAP thresholds than CI users whose electrode arrays were seated more laterally in the cochlear duct (e.g. the CI24M and CI422 implants). The position of the electrode contacts relative to the modiolus of the cochlea was found to be related to slope of the ECAP growth functions. The lowest slopes were found in CI24RE users. It also had a significant impact on the width of the channel interaction function. Electrode arrays seated further from the modiolus have significantly more channel interaction than electrode arrays that hug the modiolus of the cochlea. Differences between results recorded from post-lingually deafened adults and children with congenital hearing loss were minimal. The difference only reflected on the ECAP slopes. Slopes in children with congenital hearing loss were significantly steeper than those recorded from adults. This may indicate that children with congenital hearing loss may have better neural survival than adults with acquired hearing loss. In conclusion, the results of the current study show evidence of the effects of variations in design and function of the implanted components of the Nucleus CI. Perhaps the most significant finding from the current data set is that electrode arrays located closer to the modiolus of the cochlea have lower thresholds and exhibit less channel interaction than electrode arrays that are positioned more laterally. An argument could be made that lower stimulation levels and less channel interaction may result in better outcomes and/or longer battery life. For CI candidates who do not have significant residual acoustic hearing, the CI24RE implant might be a better choice than the more recently introduced CI422 electrode array.

Effective cochlear implant fitting (or programming) is essential for providing good hearing outcomes, yet it is a subjective and error-prone task. The initial objective of this research was to automate the procedure using the auditory nerve electrically evoked compound action potential (the ECAP) and machine intelligence. The Nucleus?? cochlear implant measures the ECAP via its Neural Response Telemetry (NRT™) system. AutoNRT™, a commercial intelligent system that measures ECAP thresholds with the Nucleus Freedom™ implant, was firstly developed in this research. AutoNRT uses decision tree expert systems that automatically recognise ECAPs. The algorithm approaches threshold from lower stimulus levels, ensuring recipient safety during postoperative measurements. Clinical studies have demonstrated success on approximately 95% of electrodes, measured with the same efficacy as a human expert. NRT features other than ECAP threshold, such as the ECAP recovery function, could not be measured with similar success rates, precluding further automation and loudness prediction from data mining results. Despite this outcome, a better application of the ECAP threshold profile towards fitting was established. Since C-level profiles (the contour of maximum acceptable stimulus levels across the implant array) were observed to be flatter than T-level profiles (the contour of minimum audibility), a flattening of the ECAP threshold profile was adopted when applied as a fitting profile at higher stimulus levels. Clinical benefits of this profile scaling technique were demonstrated in a 42 subject study. Data mining results also provided an insight into the ECAP recovery function and refractoriness. It is argued that the ECAP recovery function is heavily influenced by the size of the recruited neural population, with evidence gathered from a computational model of the cat auditory nerve and NRT measurements with 21 human subjects. Slower ECAP recovery, at equal loudness, is a consequence of greater neural recruitment leading to lower mean spike probabilities. This view can explain the counterintuitive association between slower ECAP recovery and greater temporal responsiveness to increasing stimulation rate. This thesis presents the first attempt at achieving completely automated cochlear implant fitting via machine intelligence; a future generation implant, capable of high fidelity auditory system measurements, may realise the ultimate objective.

This study determined whether the Hodgkin-Huxley model for unmyelinated nerve fibres could be more comprehensively modified to predict excitation behaviour at Ranvier nodes of a human sensory nerve fibre, as specifically applied to the prediction of temporal characteristics of the human auditory system. The model was developed in three phases. Firstly, the Hodgkin-Huxley model was modified to describe action potential dynamics at Ranvier nodes using recorded ionic membrane current data from single human myelinated peripheral nerve fibres. A nerve fibre cable model, based on a combination of two existing models, was subsequently developed using human sensory nerve fibre morphometric data. Lastly the morphological parameters of the nerve fibre model were changed to resemble a Type I peripheral auditory nerve fibre and coupled to a volume-conduction model of the cochlea. This study is the first to show that the Hodgkin-Huxley model equations can be modified successfully to predict excitation behaviour of a generalised human peripheral sensory nerve fibre without using the Goldman-Hodgkin-Katz equations. The model includes a more comprehensive establishment of temperature dependence of the physiological and electrical parameters compared to existing models. Two versions of the human Type I auditory nerve fibre model were developed, one simulating an undamaged (non-degenerate) fibre and another a damaged (degenerate) fibre. Comparison between predicted and measured results indicated similar transient and persistent sodium, as well as slow potassium ionic membrane currents to those found in generalised sensory nerve fibres. Results confirm that chronaxie, rheobase current, mean latency, threshold and relative refractive periods depend on the amount of degeneracy of fibres. The model could account for threshold differences observed between different asymmetric waveforms. The combination of persistent sodium and slow potassium ionic membrane currents could in part predict non-monotonic excitation behaviour observed experimentally. A simplified method was developed to calculate electrically evoked compound action potential responses following neural excitation. It provided a computationally effective way to obtain an estimate of profile widths from the output of models that calculate neural excitation profiles, and an indirect way to estimate stimulus attenuation by calculating the value of the parameter that produces the best fit to experimental data. Results also confirmed that electrode arrays located closer to the modiolus produce more focussed neural excitation spread than more laterally located arrays. / Thesis (PhD)--University of Pretoria, 2010. / Electrical, Electronic and Computer Engineering / unrestricted

<div>Electrical nerve stimulation (ENS) is an emerging therapy for many neurological disorders. Compared with conventional one-way stimulations, closed-loop ENS approaches increase the stimulation efficacy and minimize patient's discomfort by constantly adjusting the stimulation parameters according to the feedback biomarkers from patients. Wireless neurostimulation devices capable of both stimulation and telemetry of recorded physiological signals are welcome for closed-loop ENS systems to improve the quality and reduce the costs of treatments, and real-time digital signal processing (DSP) engines processing and extracting features from recorded signals can reduce the data transmission rate and the resulting power consumption of wireless devices. Electrically-evoked compound action potential (ECAP) is an objective measure of nerve activity and has been used as the feedback biomarker in closed-loop ENS systems including neural response telemetry (NRT) systems and a newly proposed autonomous nerve control (ANC) platform. It's desirable to design a DSP engine for real-time processing of ECAP in closed-loop ENS systems. </div><div><br></div><div>This thesis focuses on developing the DSP architecture for real-time processing of ECAP, including stimulus artifact rejection (SAR), denoising, and extraction of nerve fiber responses as biomedical features, and its VLSI implementation for optimal hardware costs. The first part presents the DSP architecture for real-time SAR and denoising of ECAP in NRT systems. A bidirectional-filtered coherent averaging (BFCA) method is proposed, which enables the configurable linear-phase filter to be realized hardware efficiently for distortion-free filtering of ECAPs and can be easily combined with the alternating-polarity (AP) stimulation method for SAR. Design techniques including folded-IIR filter and division-free averaging are incorporated to reduce the computation cost. The second part presents the fiber-response extraction engine (FREE), a dedicated DSP engine for nerve activation control in the ANC platform. FREE employs the DSP architecture of the BFCA method combined with the AP stimulation, and the architecture of computationally efficient peak detection and classification algorithms for fiber response extraction from ECAP. FREE is mapped onto a custom-made and battery-powered wearable wireless device incorporating a low-power FPGA, a Bluetooth transceiver, a stimulation and recording analog front-end and a power-management unit. In comparison with previous software-based signal processing, FREE not only reduces the data rate of wireless devices but also improves the precision of fiber response classification in noisy environments, which contributes to the construction of high-accuracy nerve activation profile in the ANC platform. An application-specific integrated circuit (ASIC) version of FREE is implemented in 180-nm CMOS technology, with total chip area and core power consumption of 19.98 mm² and 1.95 mW, respectively. </div><div><br></div>