Stroke can affect all levels of the auditory system (from the inner ear to the central tracts), and may result in various types of auditory dysfunctions, such as peripheral hearing loss (cochlea to auditory nerve), disordered auditory processing (brainstem to cortex), and cortical deafness. Hearing-impaired stroke survivors have an increased risk of physical decline after discharge to the community. This may be attributed to restricted participation in post-acute rehabilitation programs due to the hearing impairment. Furthermore, hearing impairment may have a significant impact on listening, linguistic skills and the overall communication of the affected stroke patient. To date, no studies have sought to systematically characterise the auditory function of stroke patients in detail in order to establish the different types of hearing impairments in this cohort of patients. Such information would be clinically useful for understanding and addressing the hearing needs of stroke survivors so that appropriate management could be given to these patients in order to improve their quality of life. One of the main aims of this research was to characterise and classify the hearing impairments of stroke patients using a detailed audiological assessment test battery in order to determine the level of clinical need and inform appropriate rehabilitation for this patient population. We found that the most common type of hearing impairment in stroke subjects was the combination type, ‘peripheral hearing loss and central auditory processing disorders’, in the older subgroup (in 55%), and auditory processing deficits in the younger subgroup (in 40%). Both types of impairment were significantly higher in these groups than in controls. Offering a comprehensive audiological assessment to all stroke patients would be a costly and time-consuming process. Therefore, a preliminary screening program for such patients needs to be identified, e.g. by means of a questionnaire, so that the full audiological assessment could be reserved for those who fail the initial screening. We aimed to determine whether a handheld hearing screener together with two validated hearing questionnaires could be used as a hearing screening tool to facilitate early identification and appropriate referral of hearing impaired stroke patients in the subacute stage. The highest test accuracy was achieved when results of the handheld hearing screener and hearing questionnaires were combined. Nehzat Koohi PhD Thesis vi Auditory disability due to impaired auditory processing (AP), despite normal puretone thresholds, is common after stroke. However, there are currently no proven remedial interventions for AP deficits in stroke patients. Our study is first to investigate the benefits of personal frequency-modulated (FM) systems in stroke patients with disordered AP. Our results demonstrated that FM systems may substantially improve speech-in-noise deficits in stroke patients who are not eligible for conventional hearing aids. We also evaluated the long term benefits for speech reception in noise, after daily ten-week use of personal FMs, in non-aphasic stroke patients with auditory processing deficits. We found that ten weeks of using FM systems by adult stroke patients may lead to benefits in unaided speech in noise perception. Our findings may indicate auditory plasticity type changes.

How can we understand speech in difficult listening conditions? This question, centered on the ‘cocktail party problem’, has been studied for decades with psychophysical, physiological and modelling studies, but the answer remains elusive. In the cochlea, sounds are processed through a filter bank which separates them in frequency bands that are then sensed through different sensory neurons. All the sounds coming from a single source must be combined together again in the brain to create a unified speech percept. One of the strategies to achieve this grouping is to use common sound source location. The location of sound sources in the frequency range of human speech in the azimuthal plane is mainly perceived through interaural time differences (ITDs). We studied the mechanisms of ITD processing by comparing vowel discrimination performance in noise with coherent or incoherent ITDs across auditory filters. We showed that coherent ITD cues within one auditory filter were necessary for human subjects to take advantage of spatial unmasking, but that one sound source could have different ITDs across auditory filters. We showed that these psychophysical results are best represented in the gerbil inferior colliculus when using large neuronal populations optimized for natural spatial unmasking to discriminate the vowels in all the spatial conditions. Our results establish a parallel between human behavior and neuronal computations in the IC, highlighting the potential importance of the IC for discriminating sounds in complex spatial environments.

Introduction Rates of surgery vary considerably across the UK. Many assume that this on the one hand exposes residents of certain UK regions to unnecessary surgical risks, and on the other hand prevents those of neighbouring regions from receiving important surgical care. Nowhere is this more apparent than for tonsillectomy – an operation that involves removing the tonsils in patients suffering with severe recurring sore throats. With 40,000 tonsillectomies per year, it is one of the most common operations in the NHS, but remarkably, tonsillectomies are done seven times more frequently in some UK regions than in others. Despite national efforts to reduce these differences (e.g. re-education programmes for ENT surgeons, creation of national guidance on how to manage recurring sore throat, and financial penalties locally imposed to restrict high numbers of surgeries) this disparity has only got worse over the past 17 years. I undertook my research to examine the causes of these differences in more detail, to guide future policies. Aims The aim of my thesis was to develop a better understanding of the drivers of regional tonsillectomy rate variation by quantifying regional variation of tonsillectomy rates in relation to regional demands, and by exploring the role of professional uncertainty and treatment preference on the treatment chosen. Objectives The objectives were to establish the: A. Rate and regional variation of self-reported sore throat and help seeking behaviour in the community; B. Rate and regional variation of recurring sore throat in primary care; C. Rate and regional variation of tonsillectomy in secondary care, after adjusting for local rates of recurring sore throat; D. Constructs of clinical decision making and thereby ascertain which concepts were most likely to be related to surgical rate variation; E. Role of surgeon and patient decisional uncertainty on the treatment chosen for recurring tonsillitis; F. Role of surgeon and patient treatment preference on the treatment chosen for recurring tonsillitis; Methods By using the largest UK population based study of upper respiratory symptoms and primary-secondary care linked medical record databases, I was able to investigate regional surgical rate variations across the entire patient-disease pathway: from sore throat in the community (used as a surrogate marker for tonsillitis), through recurrent sore throat consultations in primary care and finally tonsillectomy in hospitals. Following systematic review and thematic analysis of patient focus groups, I designed an instrument to elicit patient and surgeon preference. By undertaking the largest observational study of decision-making in adults with recurring tonsillitis, I was able to investigate the role of both patient and surgeon treatment preference and decisional uncertainty on treatment choice. Results My results suggest: A. There is considerable variation in the incidence of self-reported severe sore throat between regions. However, once patient risk factors are accounted for there is no statistical evidence for disparity between regions. In those who self-report a severe sore throat there is also a degree of regional variation in the rate of relevant consultations for sore throat symptoms, however, once disease characteristics were accounted for, this regional disparity disappears. B. There is regional disparity in recurring sore throat consultations in primary care, however, once patient characteristics are accounted for, this regional variation reduces considerably. C. Similarly, there is regional disparity in tonsillectomy rates; this variation reduces considerably once patient characteristics are accounted for. D. In the literature concepts related to shared-decision-making are strongly inter-related and often poorly defined. Decisional uncertainty and treatment preference are amongst the best described, most measurable, and most appropriate constructs to investigate in a study of surgical rate variation. E. Decisional uncertainty, either patient’s or surgeon’s, was found to have no role to play in the treatment chosen during a consultation for recurring sore throat. F. Patients’ treatment preferences did not influence their treatment chosen, but surgeons’ treatment preferences did. Discussion There are three key findings of my thesis. Firstly, regional rate of consultation for sore throat – which was used as a proxy for the underlying tonsillitis rate in the population throughout – was mirrored in the regional rate of tonsillectomy. This implies the regional tonsillectomy rate variations reflect regional variations in the ‘need’ of the population. Secondly, regional tonsillectomy rate variations are greater for children than adults. Finally, treatment decisions for adults with recurring tonsillitis are more influenced by surgeon’s treatment preferences than patient preferences or severity. There is a strong culture within the NHS of addressing variations of all kinds as a means of increasing healthcare quality and decreasing cost. There are currently metrics of variation across almost every aspect of care, however few of these account for patient characteristics to the extent that this thesis has, meaning that the initiatives may be a waste of effort at best and harmful at worst. The work presented in my thesis uses a unique set of mixed methods to demonstrate the complexity of regional tonsillectomy rate variation, which too frequently has been investigated using poorly controlled cross-sectional studies and reduced to soundbites like the “Surgical Signature”. Whilst my study shows “surgical signature” is important, it fails to describe the true complexity of the variation observed. My study sheds more light on the complexity of this variation and provides a plausible reason as to why the policies to reduce tonsillectomy rate variations may have failed. This mixed methods approach could be used more broadly to inform discussions under regional surgical rate variations. Most importantly, the findings in this thesis also demonstrate where future policy could be targeted to reduce unwarranted regional tonsillectomy rate variation.

Type I spiral ganglion neurons (SGNs) synapse onto cochlear inner hair cells and constitute the majority of afferent fibres in the auditory nerve (AN). Better characterisation of their biophysical properties may identify therapeutic targets for optimising AN sensitivity. This study aimed to characterise the membrane physiology underlying the firing properties of post-hearing onset SGNs and investigated whether their properties could be modified by the presence of native and synthetic lipids. In dissociated ganglionic cultures, SGNs displayed an intrinsic variation in their firing properties; this could be correlated with the magnitudes of specific membrane currents. SGNs were categorised by their response to depolarising current injection; SGNs either adapted to the stimulus rapidly, slowly or not at all. Rapid adaptation, a mechanism that preserves temporal precision throughout the auditory system, was found to be regulated by a dendrotoxin-K (DTX-K) and tityustoxin-Kα (TsTx)-sensitive low-threshold voltage-activated (LVA) K+ current, suggesting contribution by Kv1.1 and Kv1.2 subunits. As Kv1.2 channels were known to be positively modulated by membrane phosphoinositides, we investigated the influence of phosphatidylinositol-4,5- bisphosphate (PIP2) availability on SGN K+ currents. Inhibiting PIP2 production using wortmannin, or sequestration using a palmitoylated peptide (PIP2-PP), slowed or abolished adaptation in SGNs. PIP2-PP specifically reduced SGN LVA currents in a manner that was partly rescued by intracellular dialysis with diC8PIP2, a nonhydrolysable analogue of PIP2. PIP2-PP application induced similar levels of current inhibition in Kv1.1/Kv1.2 channels heterologously expressed in HEK293 cells. Accordingly, the lipid sensitivity of the Kv1.2 channel was further explored with a range of native and synthetic free fatty acids. Polyunsaturated fatty acids were found to be strong inhibitors of Kv1.2 currents, offering further potential candidates for SGN modulation. Collectively, this data identifies Kv1.1 and Kv1.2 containing K+ channels as key regulators of excitability in the AN, and potential targets for pharmacological modulation.

The unifying theme of this thesis is a series of research studies that collectively amount to a feasibility study for clinical trials of laryngeal reinnervation for the treatment of vocal fold paralysis. The question ‘Does laryngeal reinnervation or thyroplasty give better voice results for patients with unilateral vocal fold paralysis (UVFP)?’ remains outstanding; a question that ideally requires a randomised control trial. However, randomised control trials in surgery face inherent surgeons’ equipoise and recruitment issues that may lead to its failure. I performed a national survey of UK ENT consultants exploring their perception and obtaining crude numbers of eligible UVFP patients under their care for such trial, which revealed that the majority of ENT surgeons are receptive to the trial and the size of the potential patient pool is promising. I interviewed eligible UVFP patients to explore issues around the recruitment process, and this suggested that the proposed trial is feasible. Some phraseology used during recruitment that needed changing was identified, which may optimise the recruitment process for a trial. In voice surgery trials, outcome measures should be multidimensional and standardised. Acoustic analysis has been proposed but has limitations. OperaVOX is a potential new acoustic analysis software developed to resolve some of these factors. I demonstrated that OperaVOX is statistically comparable to the ‘gold standard’, Multidimensional Voice Programme, for most principal phonatory outcome measures. Another outcome measure- video-laryngostroboscopy, allows visual evaluation of characteristics and vibratory pattern of vocal folds. It is typically subjective that requires inter- and intra-rater reliability study. Here, I demonstrated that certain parameters depicted substantial inter- and intra-rater reliability. However, I showed that rater training is required to improve the reliability of other parameters. I investigated MRI as a potential non-invasive method to evaluate vocal muscles’ denervation and reinnervation. I found that signal changes on the T2-weighted MRI larynx images correlated with electrophysiological results with good repeatability. Another MRI sequence, dynamic contrast enhanced- and diffusion weighted MRI, suggested reduced perfusion in paralysed muscles, whilst cine-MRI for vocal fold mobility assessment demonstrated considerable potential as a method to grade vocal fold mobility. Finally, I present a small prospective case series of non-selective and selective laryngeal reinnervation in UVFP and unilateral vagal paralysis following vagal tumour excision respectively concomitant with injection laryngoplasty. Voice improvement was demonstrated by voice handicap index-10 and other multidimensional outcome measures, and these were supported by laryngeal electromyography and T2-weighted MRI outcomes. To my knowledge, this is the first multidimensional prospective study of laryngeal reinnervation and also the first to suggest that 3T MRI may be a promising outcome measure for future reinnervation trials. In summary, I have shown that a randomised trial of laryngeal reinnervation versus thyroplasty is feasible in the UK, and have validated patient- and observer-rated outcome measures. I have also shown that MRI may offer an alternative to electromyography in the assessment of laryngeal neuromuscular function in future trials and the clinic.

Aims and objectives: The overall aim of this thesis is to investigate the neural substrates of tinnitus perception using advanced magnetic resonance imaging (MRI) techniques. The main objectives of this thesis dissertation are to (1) identify the impact of tinnitus perception on the quality of life, (2) investigate the morphological alterations in cortical and subcortical brain structures in tinnitus, (3) explore the auditory perception in tinnitus, (4) identify the perfusion pattern changes in tinnitus, (5) investigate the effect of tinnitus perception on brain functional connectivity, (6) explore the microstructure alterations in white matter structures in tinnitus and (7) investigate the relationship between tinnitus handicap scores and brain structure/function. Materials and methods: A total of 34 individuals with tinnitus, 20 healthy controls with mild to moderate hearing loss (MH), and 20 healthy controls with normal hearing (NH) participated in the work presented in this thesis. Pure tone air conduction audiometry was performed to assess the hearing level. Behavior assessments were undertaken of handedness, anxiety and depression, and tinnitus severity. Different MR images were acquired: T1-weighted images, T2-weighted images, functional images (resting-state and task-based fMRI), arterial spin labelling images (ASL) and diffusion tensor imaging (DTI). Different MRI analysis techniques were applied including: voxel and surface based morphometry (VBM and SBM), shape appearance differences, independent component analysis (ICA), and tractbased spatial statistics (TBSS).

Otitis media is one of the most common childhood ailments seen in clinical practice today, yet the condition has rarely been identified in the skeletal remains of past populations. Other diseases of the middle ear and mastoid have been sporadically reported in the archaeological literature but no standardised criteria for their diagnosis in temporal bones have been established. A comparative study of different non-destructive methods of examination of human temporal bones was undertaken in order to determine the optimum strategy for future research. The material for the study consisted of 1244 temporal bones representing 688 individuals from seven English cemeteries dated to the Romano-British, Anglo-Saxon and medieval periods. The examination of such a wide range of material facilitated determination of optimum diagnostic criteria while eliminating single site bias. It also allowed a comparison to be made, of the prevalence of ear disease, between populations separated both geographically and chronologically. The inclusion in the study of the human skeletal meterial from a medieval leprosy hospital cemetery at Chichester afforded the opportunity to investigate the possibility that ear disease was more prevalent in those individuals who had suffered from leprosy. A review of the documentary evidence for the treatment of middle ear disease in the past suggests that, prior to the introduction of antibiotic therapy, no truly effective treatment was available and, therefore, palaeopathological evidence is unlikely to have been influenced by therapeutic intervention. In the preparation of skeletal reports the temporal bones should, ideally, be examined using a combination of methods. Two methods are recommended as being the most useful where resources are limited but in the identification of particular pathological conditions other methods proved useful. This research has demonstrated that middle ear disease in ancient populations can, with some certainty, be diagnosed using the proposed criteria, thus removing the barrier to future comparative research in this field.

Acoustic over exposure (AOE) triggers hearing loss and tinnitus but cellular mechanisms underlying those auditory defects are still poorly understood. This thesis explores the changes of excitability produced by AOE in identified cells of the rat dorsal cochlear nucleus (DCN) within the auditory brainstem. A development of a method combining Golgi silver impregnation with Nissl staining allowed study of the morphology and the distribution of the main DCN neuronal subtypes within slices containing the DCN. Whole cell patch clamp recordings allowed characterisation of the distinctive electrophysiological properties of the main DCN neuronal subtypes. In vitro stimulations of auditory or multisensory synaptic inputs showed fundamental differences in terms of the principal neurones firing pattern and the role of inhibitory synaptic transmission on firing pattern. Wistar rats were exposed to loud (110 dB SPL) single tones (15 kHz) for a period of 4 hours (protocol of AOE). Non invasive auditory brainstem response recordings were performed after 3 to 4 days and showed a significant increase of the rat’s hearing threshold for frequencies above 8 kHz. Whole cell recordings performed at a similar time (3 to 4 days) after AOE, showed that AOE led to a change of the passive and the active properties of DCN interneurones and principal cells. AOE also decreased the general excitability of the cellular network and affected differently excitatory and inhibitory synaptic transmission onto principal neurones depending on whether multisensory or auditory synaptic inputs were stimulated. Computational modelling allowed simulation of the effects of AOE on principal cell firing patterns and elaboration of a general theory whereby AOE triggers shifts of hearing thresholds concomitant with plastic adjustments in the DCN network. In conclusion, an elevation of the hearing threshold accompanied by significant excitability changes within the central auditory system could represent fundamental steps towards the development of tinnitus.

Substantial evidences show that tinnitus is associated with hearing loss. However, approximately 10% of tinnitus patients attending an ENT/Audiology clinic for their problems have hearing within normal audiometric thresholds. The general consensus on the roles of hearing loss in triggering tinnitus seems not applicable in this minority group. The absence of any grossly abnormal audiometric findings to explain their tinnitus leaves the clinician with a diagnostic dilemma and presents difficulties in dealing with the patients. The aim of this study was therefore to scrutinize auditory functioning in a sample of tinnitus subjects with normal hearing thresholds and non-tinnitus normally hearing control participants. Five tests were applied: Audioscan, Transient Evoked Otoacoustic Emission (TEOAE), Distortion Product Otoacoustic Emission (DPOAE), Threshold Equalising Noise (TEN) test and Contralateral Suppression of TEOAE (CSTEOAE). Optimal parameters for these tests have been defined and subsequently used to detect subtle auditory deficit in tinnitus patients with normal hearing thresholds. Twenty-seven tinnitus patients with hearing thresholds within normal limit participated in this study. A total of 27 non-tinnitus volunteers with similar thresholds were recruited as controls. Significant more Audioscan notches were found in tinnitus patients than in controls, particularly at 3001 to 8000 Hz region. Moreover, abnormal TEOAE was significantly more in tinnitus patients than in controls. No significant difference was observed in DPOAE, TEN and CSTEOAE test. Only one patient was found with abnormal TEN test and no TEOAE suppression, which led to the discovery of a large vestibular schwannoma. VI In summary, these results confirm the possible existence of subtle auditory deficit in tinnitus patients with normal hearing thresholds. However, combination of cochlear and retrocochlear abnormality may also possible. The presence of tinnitus without concomitant hearing loss in this group suggests that it may be an early sign of diseases that are only diagnosed after the onset of hearing loss.

High levels of distress are often reported among people with Meniere's disease (MD). The aim of the research programme was to identify modifiable psychological factors that influence adjustment, to inform future support for people with MD. Part of the research programme was nested within a randomised controlled trial assessing the effectiveness of physical or psychological based self-treatment. Expectations and beliefs about illness and intolerance of uncertainty assessed at baseline were found to be correlated with baseline anxiety, and predicted adjustment outcomes at the end of the 3 month self-treatment period. . Few studies have considered psychological mechanisms that might explain MD related distress. Therefore a systematic review on the role of psychological factors in MD was carried out and examined the literature for the possible presence of components of four mechanisms of distress (post-traumatic stress disorder (PTSD), worry, health anxiety and anxiety sensitivity). The most evidence was found for the possible presence of PTSD and health anxiety. These mechanisms were therefore measured explicitly. A greater proportion of participants with MD had clinical levels of distress on measures ofanxiety, depression and PTSD than in the general population. Participants with MD were also significantly more distressed than healthy controls in terms of anxiety, depression, health anxiety, and health anxiety related beliefs about the negative consequences of illness. Adjustment to MD appears to be influenced by expectations and beliefs about illness, intolerance of uncertainty, PTSD, and health anxiety in different ways. With further development of empirically sound research including more longitudinal and qualitative research, a greater understanding of the mechanisms linking MD with adjustment may enable psychological treatment and support to be more effectively tailored to the particular problems ofpeople with MD.