Postural instability is one of the cardinal symptoms of Parkinson’s disease (PD). Postural instability can present on diagnosis, and commonly becomes more prominent with disease progression, resulting in subsequent falls and diminished quality of life. The treatment of postural instability is challenging, as it is often refractory to management with levodopa and deep brain stimulation of conventional targets such as the subthalamic nucleus. To assess postural instability, the most commonly used measure in the clinical setting is the pull test according to item 30 of the Unified Parkinson’s disease rating scale (UPDRS), where an examiner performs a brisk backward tug at the patient’s shoulder level and grades the corrective response. While easy to administer, outcomes can vary due to variability in test administration and interpretation. A comprehensive literature review revealed laboratory based assessments provided a more objective method to measure postural responses compared to clinical assessments in people with PD. These techniques were conventionally employed in people with PD in later disease stages who already demonstrate postural instability. Laboratory based assessments presented a method to identify abnormalities before postural instability is clinically evident and effects of therapies. The recent development of instrumentation of clinical balance tests offered an alternative technique to precisely quantify postural responses. Here, we developed an instrumented version of the pull test and investigate its utility to quantify postural instability in people with PD ranging from mild to moderate disease severity. In Study 1, the sensitivity of the instrumented pull test was investigated in healthy young participants. Postural responses were modified by presenting a startling auditory stimulus concurrent with the backwards pull. Such stimuli evoke StartReact effects and are known to speed reaction times. The instrumented pull test could detect small 10 ms decreases in postural reaction time evoked by the startling stimulus. The ability to detect such changes in healthy individuals highlights the utility of instrumented techniques and justifies further investigation in people where changes to balance is of interest. Subsequently, the instrumented pull test was used to characterise postural responses in eighteen people with mild PD (Hoehn and Yahr less than 2) in Study 2. Subclinical abnormalities in trunk and step responses were detected in participants with mild PD compared to healthy controls. Furthermore, levodopa did not restore postural responses in participants with PD to that of healthy controls (Study 3). These findings demonstrate changes to postural stability can occur in mild disease. Abnormalities of postural responses which remain refractory to levodopa also suggest non-dopaminergic pathways may be implicated in the pathophysiology of postural instability in mild PD. Pedunculopontine deep brain stimulation (PPN DBS) is a therapy developed specifically to alleviate axial symptoms of gait and postural abnormalities unresponsive to conventional therapies such as levodopa. In Study 4, the instrumented pull test was used to quantify postural responses in five people with PD and moderate to severe postural instability receiving PPN DBS. Off and on stimulation, the instrumented pull test was able to detect postural responses with greater resolution compared to clinical assessments (axial items 27 to 30 of the motor subsection of the Unified Parkinson’s disease rating scale (UPDRS) and the Mini-BESTest. However, the use of the instrumented pull test, and interpretation of findings was limited by the small sample size and highly variable postural responses in participants with moderate to severe postural instability. On stimulation, improvement in overall balance scores was demonstrated across all participants with the Mini-BESTest but not axial items of the UPDRS. This thesis demonstrated the utility of the instrumented pull test as a potential assessment tool to evaluate postural instability in PD. Identification of postural abnormalities provides valuable insights in the assessment and management of postural instability in people with PD. Clinicians should consider that subclinical postural abnormalities can be present in people with mild PD, even when patients are on levodopa. Findings from this thesis strongly support the need for further studies to explore variables of postural responses that may be useful to detect people with PD at risk of falls and for clinicians to deliver targeted interventions earlier in disease course.

Retinal prostheses are a promising technology aimed at restoring vision in people suffering severe retinal degenerative conditions such as retinitis pigmentosa (RP). Present-generation devices achieve this by electrically stimulating the residual neuronal population in the retina following degeneration in order to elicit the perception of light. At present, patients implanted with these devices are able to perceive multiple localised flashes of light, termed phosphenes, which are used to build up an artificial image of the patient's surroundings. However, present-generation retinal implants lack the spatial resolution to provide a suitable replacement for everyday visual tasks. While adequate for rudimentary tasks such as object recognition, motion detection, and pattern recognition, more complex tasks such as reading, facial recognition, and independent navigation are still not possible with modern prosthetic vision devices. Two major issues that affect retinal prostheses are: the large spread of electrical potential in the retina resulting in widespread activation of neurons, undesirable electrical field interaction and the elicitation of large phosphenes that patients find difficult to discriminate between; and that many devices are not able to elicit enough phosphenes to convey complex visual information to patients. The studies presented in this thesis investigated the effectiveness of two multichannel electrical field shaping techniques: focused multipolar (FMP) stimulation and virtual electrode (VE) current steering. These techniques have shown considerable promise in studies conducted with one-dimensional neural prosthetic devices, such as cochlear and deep brain implants, as ways to restrict and `steer' electrical fields. In an effort to find new ways of improving spatial resolution, I have investigated whether these techniques can be adapted for use in a 2D retinal prosthesis. Using a normally-sighted cat model I have demonstrated that FMP stimulation is capable of restricting current spread in two dimensions and eliciting retinal and cortical response patterns with reduced spread compared with responses to the more conventional MP means of stimulation. I have also demonstrated that VE current steering between up to six electrodes can produce cortical activation patterns in predictable locations, with similar spread of neural activation as response patterns to physical electrode stimulation. By varying the proportions of charge applied to steering electrodes, it was also possible to shift the location of cortical activation in two dimensions in a predictable and intuitive fashion. To investigate the effectiveness of these techniques in a model more representative of patients, FMP stimulation and VE current steering were re-evaluated using a cat model of retinal degeneration. Unfortunately, many of the promising results from the normally-sighted cohort were not maintained when applied to degenerate retinae. While FMP stimulation still activated a localised population of retinal neurons, it was not found to elicit cortical response patterns with reduced spread compared to monopolar stimulation. The location of cortical response patterns elicited by VE stimulation were also found to be unpredictable. These results also show evidence of compressed retinotopy and increased spatial selectivity in the degenerate visual system, which significantly altered neural responses to electrical stimulation. These findings demonstrate that FMP stimulation and VE current steering, in their present form, may not be as effective in focusing and steering neural activation when applied to degenerate retinae. These results also provide a greater understanding of the differences between the responses of healthy and degenerate visual systems to electrical stimulation, which I hope will inform the further development and optimisation of these stimulation strategies.

Over recent decades, there is increasing interest in implantable devices that interact with neural tissue in the human body. Applications are broad, ranging from cardiac pacemakers to cochlear implants and beyond. The emergence of microelectronics and microfabrication has led to the miniaturization of these neural implants. Small devices are safer to implant but a number of challenges need to be addressed before very small devices are routinely deployed. For instance, it is difficult to transfer sufficient power to small implants wirelessly, and difficult to fabricate small neurostimulation microelectrodes with high enough charge injection capacity to operate safely. Compounding this, the immune system of the body can react to the implant. Unfavourable interactions of the electrode with tissue/neurons leads to a sharp drop in performance caused by scar tissue surrounding them. These challenges, among others, must be overcome in order to reduce the size of implants into the low millimeter dimensions. Devices at this scale will be insertable with minimal trauma and will hence be deployable in a greater range of circumstances. Here we investigate the use of diamond as a biomaterial with the potential to mitigate or ameliorate some of these challenges. In this work, a novel technique for microcoil fabrication is introduced. Trenches were milled into a diamond substrate and filled with silver active braze alloy, enabling the manufacture of small, high cross-section, low impedance microcoils capable of wireless power transmission of 10 mW over 6 mm. The coils were encapsulated in a second layer of diamond, characterized, and accelerated ageing was performed to verify the longevity of the construct. Building on previous work, a method was developed to grow conducting diamond films on platinum foil. A laser roughening method was used to increase adhesion of the diamond to the platinum. This approach enables the superior properties of diamond to be integrated into devices constructed using traditional fabrication methods such as wire bonding or laser welding. Laser roughened platinum was coated with nitrogen induced ultra-nanocrystalline diamond (N-UNCD) films and the electrochemical performance of these films was measured relative to platinum. Stronger attachment of N-UNCD to platinum substrates of higher roughness was observed. Diamond on platinum electrodes were found to be more capacitive and stable compared to platinum controls, a favorable characteristic for neural stimulation. Finally, an extracellular matrix protein (laminin) known to be involved in inter-neuron adhesion and recognition, was covalently coupled to diamond electrodes. Biologically, active interlayers have the potential to increase neural adhesion to electrodes and/or reduce the immune response, thus increasing longevity. Electrochemical analysis found that covalently coupled films were robust and resulted in minimal change to electrochemical properties of the electrodes. Neurons cultured on laminin coated surfaces exhibited improved adhesion. This thesis demonstrates that diamond is a versatile material for use in medical implants. It can be used as a construction material and as an encapsulant containing electrically active elements. It can be made electrically conductive and possesses suitable electrochemical properties for neural stimulation. Finally, it can be employed as a chemically active substrate for attachment of additional chemistries, including biomolecules.

While cochlear implants (CI) have been successful in restoring a sense of hearing to people with severe to profound sensorineural hearing loss, there is still a wide variance in speech outcomes for CI users. Psychophysical experiments have shown that some of the variance can be explained by sensitivity to temporal modulations. If poor outcomes are partly caused by limited access to temporal speech cues, then improved transmission of those cues may provide perceptual benefits to CI users. The broad aim of the research presented in this thesis was to improve speech outcomes for CI users through better transmission of temporal speech information. The first study investigated the effect of stimulation rate and presentation level on speech perception and temporal modulation detection for CI users, in order to identify stimulation rates that provide a perceptual advantage in certain listening conditions. Speech perception (in quiet and in noise) and amplitude modulation detection thresholds (AMDTs) were measured at different presentation levels and stimulation rates. The reduction in speech perception due to added noise was significantly greater at higher rates. Speech perception was also significantly affected by presentation level, with scores increasing as the level increased. In contrast, AMDTs, as measured via acoustic input to the speech processor, exhibited no effect of rate or level. Correlations were found between AMDTs and speech perception for both the low-rate and high-rate processors. Therefore, while AMDTs explained inter-subject variability in speech perception, they did not explain within subject variability across stimulation rates and presentation levels. The second study evaluated the perception of amplitude modulation (AM) and rate modulation (RM), providing fundamental information about the temporal processing abilities of CI users and the perceptual mechanism underlying those abilities. The study was the first psychophysical evaluation of AM and RM detection in the same CI users. AM and RM detection thresholds were correlated and exhibited similar effects of modulation frequency and presentation level, indicating that AM and RM may be perceived by a common perceptual mechanism that involves central temporal integration. In the final study, a novel speech processing strategy called ARTmod (Amplitude and Rate Temporal modulation) was developed and tested. The ARTmod strategy encoded speech with simultaneous AM and RM, in order to observe whether RM can be used to enhance the perception of temporal envelopes of speech signals. In the experiment, the amount of AM was fixed and the amount of RM was varied, and speech perception was measured for the different RM amounts. A significant effect of RM amount was found, with speech scores improving as the RM amount was increased. The results indicated that RM can constructively combine with AM to enhance the perception of temporal speech envelopes and improve speech perception for CI users.

Cochlear implant (CI) users differ in their auditory speech understanding ability. This variability is partly due to variability in deafness history and pathology, and partly due to functional brain changes that are likely to occur during deafness and after implantation. By measuring cortical activity in CI users, a relation between functional changes in language associated regions of their brain and speech understanding may be revealed. However, when investigating cortical activity in CI users, commonly-used neuroimaging techniques have limitations. For example, EEG and fMRI may suffer from magnetic or electrical artefacts, and PET imaging is invasive for participants. The studies described in this thesis used a non-invasive technique – functional near-infrared spectroscopy (fNIRS) – to investigate cortical activity in CI users related to speech understanding and the integration of audio-visual speech cues. Compared to fMRI, fNIRS also has the advantages of being quiet (not suffering from the loud magnetic scanning noise) thus suitable for auditory-related tasks, and more tolerant of body movement. The first study determined whether fNIRS measures of cortical activity in post-lingually deafened CI users when listening to or watching speech are correlated with their auditory speech understanding. The fNIRS results showed that speech-evoked cortical activity in CI users that was not only different from normal-hearing listeners but also was negatively correlated with the speech understanding ability. That is, CI users who had poorer auditory speech understanding ability showed higher fNIRS activation in certain brain regions of interest when listening to or watching speech. The increased brain responses might be related to brain functional changes that occurred in CI users during deafness and after implantation for visual speech processing or more listening effort and more neural responses that were used by CI users to process auditory speech. The second study determined whether audio-visual (AV) integration of speech cues in post-lingually deafened CI users is different from that in their similar-aged normal-hearing adults. Participants’ reaction times, response accuracy, and cortical activity were measured when performing different speech identification tasks. A novel method was proposed that combined a probability model and a cue integration model to quantify the amount of AV integration based on response accuracy measures. Consistently, behavioural results using response accuracy and reaction time measures did not show better AV integration in CI users compared to people who had normal hearing. In addition, fNIRS measures of cortical activity did not show AV integration in either CI users or normal-hearing adults. The third study determined whether aging affects AV integration in people who have normal hearing when responding to speech using the same behavioural and fNIRS measures as in the second study. Again, fNIRS results did not show AV integration in either younger or older participants. Behavioural results found no significant difference in AV integration between the older and young participants using both reaction time and response accuracy measures. This thesis integrates knowledge from multisensory neuroscience and psychophysics and uses a novel brain imaging technique to measure cortical activity in CI users for language processing. Results in this thesis showed that this novel imaging technique – fNIRS – could be implemented to examine the variances in auditory speech understanding among CI users. It makes a new advance in the way that multisensory abilities are measured behaviourally, by combining models of optimal and minimum integration. Results in this thesis found that there was no significant difference between CI users and normal-hearing adults in the integration of audio-visual speech cues. Neither was there a significant effect of aging on AV integration.

Retinal prostheses are a promising treatment for blindness caused by photoreceptor degeneration. Electrodes implanted in the retina deliver electrical stimuli in the form of current pulses that activate surviving neurons to restore a sense of vision. Clinical trials for such devices have shown that the visual percepts evoked are informative, and can improve the day-to-day life of recipients. However, the spatial resolution of retinal prostheses is a limiting factor, with those who have the highest reported acuity measures still classified as legally blind. Simultaneous stimulation of multiple electrodes is a possible strategy to improve device resolution without increasing the number of physical electrodes. However, electrode interactions that occur during simultaneous stimulation are not well understood. This thesis investigates the characteristics of cortical responses to simultaneous stimulation of multiple electrodes. We formulated a quantitative model to characterise the responses of visual cortex neurons to multi-electrode stimulation of the retina to understand how simultaneous stimulation can improve resolution. Activity was recorded in the visual cortex of normally sighted, anaesthetised cats in response to temporally sparse, spatially white stimulation with 21 or 42 electrodes in the suprachoroidal space of the retina. These data were used to constrain the parameters of a linear-nonlinear model using a spike-triggered covariance technique. The recovered model accurately predicted cortical responses to arbitrary patterns of stimulation, and demonstrated that interactions between electrodes are predominantly linear. The linear filters of the model, which can be considered as weighting matrices for the effect of the stimulating electrodes on each cortical site, showed that cortical responses were topographically organised. Photoreceptor degeneration results in a number of changes in the surviving cells of the retina that can negatively impact stimulation strategies. Therefore, in the second study, we investigated the effect of multi-electrode stimulation on the degenerate retina. Characteristics of cortical responses to simultaneous stimulation of multiple electrodes were evaluated in unilaterally, chronically blind anaesthetised cats, bilaterally implanted with suprachoroidal retinal prostheses. Significant differences were found between responses to stimulation of the normally sighted and blind eyes, which may help to explain the varied perceptual observations in clinical trials with simultaneous stimulation. The success of the linear-nonlinear model in predicting responses to arbitrary patterns of stimulation indicated that it may provide a basis for optimising stimulation strategies to shape cortical activity. Therefore, we investigated the possibility of inverting the model to generate stimuli aimed at reliably altering the spatial characteristics of cortical responses. An in vivo preparation with a normally sighted, anaesthetised cat showed that the response characteristics derived by the model could be exploited to steer current and evoke predictable cortical activity. Overall, these results demonstrate that cortical responses to simultaneous stimulation of both the normal and degenerate retina are repeatable, and can be predicted by a simple linear-nonlinear model. Furthermore, the interactions between electrodes are predominantly linear, and can be harnessed to shape cortical activity through inversion of the model. The method shows promise for improving the efficacy of retinal prostheses and patient outcomes.

Multichannel cochlear implants have been well accepted as an effective and safe treatment for severe to profound sensorineural hearing loss through electrical stimulation of residual spiral ganglion neurons. However, speech intelligibility with existing cochlear implants is thought to be limited by poor spatial selectivity and interactions between channels caused by overlapping activation with contemporary stimulation strategies such as monopolar (MP) stimulation. Focused intracochlear stimulation, resulting in an increase in the number of truly independent stimulating channels available for simultaneous activation, may enable better speech and pitch recognition and also improve temporal resolution. Various current focusing stimulation strategies such as tripolar (TP) stimulation have been reported to produce sharper excitation patterns and reduced channel interactions compared to MP stimulation at the cost of higher stimulation current levels. Focused multipolar (FMP) stimulation is another such focusing technique; utilizing simultaneous stimulation of multiple channels to create focused electrical fields. FMP stimulation has been validated in a small group of cochlear implant recipients showing that focusing can be achieved, however this was at the expense of higher stimulation currents compared to MP stimulation. There have been no previous attempts to systematically compare the efficacy of FMP stimulation against TP stimulation or to determine whether factors such as neural survival and the electrode position within the cochlea would affect the performance of FMP stimulation. Controlled preclinical studies in experimental animals can reduce the possible confounding effects of neural survival in human studies. It is also important to test if FMP produces non-auditory sensations since the simultaneous nature of the stimuli would be expected to require greater charge to evoke neural responses. The primary objectives of this thesis were to determine the efficacy of FMP stimulation, compared to both MP and TP stimulation, by evaluating a) the spatial extent of neural activation b) interactions between cochlear implant channels and c) modulation sensitivity to sinusoidal amplitude-modulated pulse trains. The effects of factors such as degeneration of spiral ganglion neurons, induced by long-term deafness, and the electrode position within the cochlea on the effectiveness of FMP, TP and MP stimulation were also examined. These objectives were achieved by implanting a multichannel cochlear implant into cats and guinea pigs, and recording the neural responses in the inferior colliculus in acute electrophysiological experiments. Neural thresholds and the spread of activation along the tonotopic gradient were measured. In summary, the main results of this thesis showed that FMP and TP stimulation resulted in more restricted neural activation and reduced channel interaction compared to MP stimulation and these advantages were maintained in cochleae with significant neural degeneration. Moreover, these effects were not adversely affected by the position of the electrode array within the scala tympani. Although greater charge was required to achieve threshold levels, no evidence of ectopic stimulation of non-auditory neurons was observed with FMP or TP stimulation. Systematically varying the degree of current focusing lowered threshold levels for FMP stimulation while still maintaining a selectivity advantage. Modulation detection of MP was found to be significantly better than FMP and TP stimulation at low stimulation levels, but similar at high stimulation levels. Importantly, there was no benefit in terms of restricted neural activation, reduced channel interaction or better modulation sensitivity for FMP compared to TP stimulation. The greater spatial selectivity, reduced channel interactions and the ability to convey modulation using FMP and TP stimulation would be expected to result in improved clinical performance. The insights into current focusing described in this thesis may also be helpful in other neural prostheses such as deep brain stimulation devices and visual prostheses, when more selective stimulation is desired.

Pitch is the perceptual correlate of sound frequency and is important for using speech prosody, understanding tonal languages, and music appreciation. According to pitch perception theories, pitch is encoded by the place along the cochlea that has the maximum rate of excitation and the timing of the neural activity caused by cochlear excitation. As one of the most successful neural prosthesis, cochlear implants (CIs) have enabled most recipients to achieve good speech perception in favourable listening conditions. Perceiving a precise pitch, however, is still a challenge for many CI users. The goal of this study is to develop computational models of normal and CI hearing to advance our understanding of the mechanisms of pitch perception in both cases and to investigate the factors that affect pitch perception in electrical hearing. Based on the two possible neural codes for pitch perception, two models of pitch perception were developed: the place model and the integrated model. The models simulated a common psychophysical experiment usually referred to as pitch ranking − where subjects are asked to decide which of the two sequentially presented sounds has a higher pitch – by receiving two stimuli and generating two outputs, the higher-amplitude of which would indicate the higher-pitched stimulus. Synthesised vowels with defined pitches were the sound stimuli used in this study. An artificial neural network (ANN) constituted the core of the models. Inputs to the ANN were place pitch information for the place model and both place and temporal pitch information for the integrated model. Applying the error back-propagation algorithm, the ANN was trained on a training set of pitch pairs. The performance of the pitch perception model was measured using a previously unseen test set of pitch pairs. Place code for pitch perception was extracted from simulated auditory peripheral outputs by averaging the rate of activity in the auditory nerve (AN) over time. An acoustical and an electrical model of the auditory periphery were used to simulate the activity of the AN in normal and CI hearing, respectively. The activity of the AN was further processed through a spiking neural network (SNN) to extract the temporal code of pitch. Synaptic connections in the SNN were modified by spike-timing-dependent-plasticity (STDP) to generate pitch-related precisely-timed neural activities in the SNN output neurons. Pooled inter-spike-interval histogram (ISIH) across the SNN output neurons was found to be indicative of pitch. Validation of both pitch perception models was performed by comparing their performance with psychophysical results. The place model was applied to investigate the impact of stimulation field spread on pitch perception in CI hearing using two commercial sound processing strategies. Simulation results showed that 1) the model could replicate the performance of normal and average-performing CI listeners and 2) providing focused stimulation fields in CI hearing can be beneficial, depending on the type of sound processing strategy. The integrated model was used to explore the role of and interaction between place and temporal cues in performing simulated pitch ranking tasks. Simulation results associated with the integrated model revealed that 1) temporal cues for pitch perception compensated for missing place cues in listening conditions such as a telephone conversation where low-frequency content of the signal was suppressed and 2) new strategies with improved temporal information can improve pitch perception in CI hearing, provided temporal and place information are consistent. Although drawing general conclusions about auditory perception would eventually require psychophysical experiments, computational models of auditory perception such as this work assists in focusing human testing upon factors that demonstrate the strongest impact on the auditory performance of normal and CI listeners. This would therefore lead to the more rapid development of CI sound processing strategies.

Light is transformed into neural signals by the retina. Certain conditions, such as retinitis pigmentosa, can cause extensive degeneration of the outer retinal layers, resulting in profound vision impairment. Prosthetic devices have the potential to restore visual percepts in these patients by electrically stimulating the remaining retinal neurons. One such device, the suprachoroidal retinal prosthesis, is placed between the vascular choroid and sclera. It is currently under development and is the focus of this thesis. Safety is an important aspect of medical device design. This thesis focuses on several key aspects of suprachoroidal retinal prosthesis safety. Retinal prostheses must be designed to allow for the ability to safely remove and replace the device in the case of infection, device malfunction, or a device upgrade. This thesis explores the safety and feasibility of explanting or replacing prototype suprachoroidal electrode arrays using clinical and laboratory analysis techniques in a feline model. The results indicate that suprachoroidal electrode arrays can be safely removed or replaced with minimal damage to the retina and surrounding tissues. Furthermore, the device replacement procedures were not detrimental to the retinal response to electrical stimulation. However, careful wound closure was required to minimise post-operative complications. Another key requirement of electrically active neural prostheses is that they should not cause damage that could adversely affect the efficacy of the device. The safe stimulation levels of the platinum macroelectrodes used in the prototype suprachoroidal retinal prosthesis is unknown. In addition to the above, this thesis aims to determine the safe stimulation levels of platinum macroelectrodes using electrochemical methods in vitro and in vivo. Furthermore, methods to safely increase the safe stimulation limit by altering the stimulation waveform or by using nitrogen-doped ultra-nanocrystalline diamond as an electrode material are explored. The results indicate that electrochemically safe stimulation limits were lower than stimulation levels likely to cause histologically observable damage and that altering the stimulus waveform has the potential to increase electrochemically safe stimulation limits. Also, nitrogen-doped ultra-nanocrystalline diamond electrodes have a higher safe stimulation limit than platinum and showed no signs of degradation when stimulated in vitro. The results of this thesis have helped to ensure the safety of patients implanted with prototype suprachoroidal retinal prosthesis (ClinicalTrials.gov, NCT01603576). Furthermore, this thesis has important implications on safe suprachoroidal retinal prosthesis design and makes significant contributions towards our understanding of stimulation safety.

Neuropathic pain is a chronic health condition with a severe impact on the quality of life of affected patients. The condition is often difficult to manage and refractory to traditional pain treatment strategies such as pharmacological management, physiotherapy and psychological therapy. Peripheral nerve stimulation has been proposed as an alternative treatment with numerous successful clinical reports. Nevertheless, the systematic understanding of the underlying mechanism of action is still limited. Efficacy studies in the form of randomised controlled trials have predominantly been conducted for occipital nerve stimulation to treat various headache conditions. Without trials of a wide range of neuropathic conditions, the commercial availability of approved medical devices is limited. The overall objective of this thesis was to advance towards the development of a peripheral nerve stimulation system for a small-scale clinical trial that will be used to gain a deeper understanding of the underlying mechanisms of pain modulation. Design features of electrode arrays and new stimulation strategies were tested in order to facilitate the development of advanced clinical peripheral nerve stimulation systems. The first part of the work consisted of the development of a small, wearable neural stimulator for the use in clinical trials. Chapter 2 presents the design and characterisation of the stimulator. It was shown that safe and efficacious neural activation could be achieved and the system will be suitable for use during a short-term clinical trial of electrode arrays with a percutaneous leadwire system. In the second part, a model electrode setup was used to investigate a bipolar stimulation strategy. Chapter 3 documents an electrophysiological study on the maximisation of the therapeutic window available for stimulation. An electrode screening strategy was developed in order to increase the efficiency of intra- and post-operative testing of stimulation arrays with a large number of electrode combinations. The third part of the work focussed on the development of single-source multipolar stimulation as a novel method to perform current focussing for increased selectivity of the neural activation. Chapter 4 presents the in vitro investigation that showed that a successful reduction of voltages at electrode sites other than the centre electrode was achieved when compared to monopolar stimulation. Furthermore, a significant improvement of the voltage reduction was also found compared to tripolar and common ground stimulation. The promising results from the in vitro tests were followed by an in vivo evaluation as presented in Chapter 5. However, the focussing effects found in vitro did not translate to functional benefits in vivo for the investigated setup. Rather, increased neural activation thresholds were found resulting in potentially higher power requirements for a clinical system. Monopolar stimulation was identified as the favourable mode under the tested conditions. In conclusion, the results of this thesis suggest that a safe and reliable, tailored electrode array in combination with a monopolar stimulation strategy forms a promising system in order to progress towards the overall objective, a short-term clinical trial. This will help to gain a deeper understanding of the underlying mechanism of action of peripheral nerve stimulation for the treatment of chronic neuropathic pain.