Biofeedback is a real-time training technique that involves measuring a physiological function and conveying information back to the patient, to help them learn to adjust their performance. Biofeedback is used successfully in many areas of neuromuscular rehabilitation and sports training. Lower limb amputees particularly present a need for augmented feedback. Following amputation the proprioceptive pathways required to regulate gait are impaired. During physiotherapy patients have a limited view of their body, and the physiotherapist may not be best physically placed to witness gait changes. Outside the clinic patients often adopt poor walking patterns, such as circumduction and abduction, which can lead to lower back pain and a reduced quality of life. This work focused on the development of a biofeedback training system to assist in the reduction of habitual circumduction and abduction gait patterns seen in trans-femoral amputees. Guided by a review of the literature, a training system was developed that uses electro-tactile sensory stimulation to provide feedback of the patient's thigh motion whilst they walk on a treadmill. A greater understanding of the psychophysical response to electro-tactile stimulation was required in order to present discernible information in a safe and comfortable manner. Thirteen healthy subjects were therefore recruited into a study that found thresholds of perception and discomfort to stimulation around the thigh. The study also found that subjects were able to discriminate the location of stationary stimuli and the speed and direction of moving stimuli whilst laying supine, flexing and extending the leg, and walking on a treadmill. By correctly identifying the numbered electrode locations they demonstrated an ability to perceive spatially coded information presented to the thigh using electrical stimulation. A camera-based motion capture system was incorporated into the completed biofeedback system, and software was written to capture kinematic data in real-time. To enable the calculation of feedback stimuli, a 3-dimensional biomechanical model was constructed and the patient's hip joint angles were compared to a joint angle reference database. Kinematic event detection made it possible to deliver the electro-tactile stimuli in relation to the users gait. Four amputees tested the biofeedback system and reported positively on the experience. The subjects did not walk with a circumduction gait, so it was not possible to assess the therapeutic effects of the system. However they were able to perceive and understand the feedback stimuli, relate the information to their movement, and in some cases make positive changes to their gait. Sensation threshold levels and the ability to discriminate stimuli were also found in the amputee group to be comparable with the non-amputees. This work has potential to become integrated into prosthetic components, and can be adapted for use with a broader range of patient groups with upper and lower limb movement disorders. The analysis software has potential to be further developed to provide real-time interpretation of gait patterns.

Over recent years there has been growing focus on the potential of tissue engineering approaches to produce functional bone tissue for clinical use. This thesis highlights the potential for using magnetic particle based techniques for tissue engineering applications. This study demonstrates that magnetic micro-- and nano scale particles are well tolerated by human osteoblasts over prolonged periods of time, and can be used to stimulate calcium signalling pathways in these cells. via the application offorces to integrin rc<:eptors and the internal cytoskeleton. Furthennore, this study demonstrates for the first time that human osteoblasts express the mechanosensitive 2PK+ channel TREK-I, and investigates the potential for using magnetic particle based techniques to directly and specifically activate this class of ion channel. A model system is reported that pennits the application of mechanical forces directly to distinct regions of the TREK-l channel structure and data is presented showing the direct activation of TREK-l channels following manipulation of magnetic p~icles targeted against the extended extracellular loop region of the channel structure. In addition initial data is presented concerning the effect of static magnetic fields and targeted magnetic particle manipulation on the bacterial mechanosensitive ion channel MscL. This thesis represents the first rejX>rts of targeted mechanical stimulation of mechanosensitive ion channels with magnetic particles. This approach has huge potential applications in the field of ion channel research, and offers for the first time a method to directly investigate the structural basis ofmechanosensitivity within ion channels. Furthennore the ability to manipulate ion channel activity without the need for phannacological drugs has huge jX>tential applications in tissue engineering and the treatment of conditions and diseases caused by ion channel dysfunction

Ceramics have long been used in different biomedical applications. Examples include biodegradable porous ceramic scaffolds and bioactive dense ceramic implants and prostheses. For the fabrication of different forms of ceramics, polymer binders, organic solvents and hazardous processing aids have often been used. Some of them, e.g. acrylamide monomers used in the conventional gelcasting processing. are even toxic. There is a compelling need to develop eco-friendly processing of ceramics. In this work, an environmentally-friendly protein system - egg white protein - has been employed to fabricate porous and dense ceramics to take the advantage of the well-known foaming and gelling capabilities of egg white protein.

Methacrylate-terminated poly(dimethylsiloxane) (PDMS) macromonomers have been obtained through a flexible synthetic method which has been tailored into an efficient 1 pot, 2 stage reaction. Linear and star-shaped analogues have been synthesised and 3 molecular weights of each architecture were produced for use in subsequent curing investigations. The PDMS macromonomers were photo cured into films by exposure to UV irra- diation in the presence of a free-radical initiator. A curing time study was carried out and the rheological properties, swelling ratio, Young's modulus, surface tack and macroscopic extension of the cured films were investigated. At short irradiation times, samples were sensitive to irradiation time and an increase in exposure produced a signif- icantly more solid-like film. However, after longer irradiation times, the film properties reached a plateau and films did not become significantly more solid-like despite further irradiation. Ideal network models are proposed which support the observation that the cured film properties were strongly affected by the macromonomer molecular weight. Low molecular weight macromonomers produced densely cross-linked films which were hard, elastomeric solids. Films formed from medium molecular weight macromonomers were softer materials with tacky surfaces, and films formed from the highest molecular weight were softer still and some samples exhibited viscous flow. Introducing branching into the macromonomers decreased the irradiation time re- quired to form a cohesive film. Increasing the degree of macromonomer branching increased the solid-like nature of the film in comparison to those formed from linear species, although this did not outweigh the effect of macromonomer molecular weight. The materials investigated in this study may have potential for use as non-degradable, curable materials in medical applications and have scope for in vivo curing.

Chitosan (CS) is the deacetylated derivative of the biopolymer chitin. The attractiveness of chitosan for biomedical applications is attributed to the fact that it is biodegradable, biocompatible, non-immunogenic, non-carcinogenic and, most importantly, to the fact that it is a cationic polysaccharide. Chitosan-based hydrogels have been the focus of much of this previous work. Novel chitosan based hydrogels are continuously being produced, each new gel presenting different morphologies and physical characteristics that in turn have an effect on the biomedical application they are designed for. The scope of this thesis was to further the knowledge on chitosan based hydrogel systems, designed for biomedical applications, a number of previously unreported systems are described including a novel physically crosslinked chitosan hydrogel achieved by ionic crosslinking with adenosine tri-phosphate (A TP). This produced an array of different beads and their morphological, physical and protein drug delivery properties were extensively assessed. The results showed that - these gels behaved profoundly different to similarly ionically crosslinked systems. Utilizing a novel technique, 'microbubbling', to induce porosity into otherwise low porosity hydrogels a series of gels have been successfully produced from- a high viscosity chitosan solution. Normally this technique is only applicable to low viscosity liquids but we have shown that micro bubbles were indeed introduced into the chitosan hydrogels. The morphology of these gels has been carefully analysed. Gelation of chitosan solutions has also been achieved by utilizing a novel electrochemical process. Inducing gelation of chitosan solutions was achieved in a simple electrochemical cell. The effect of various counterions was also studied. The results show that gels were produced with profoundly different morphologies. The electrochemically induced gels showed a variety of different and controllable porosities. Such gels are believed to be potential candidates for such biomedical applications such as drug delivery and tissue engineering. Chitosan also has the capability of forming 'thermosetting' gels. This thesis reports for the first time on the use of such a system as 'fiducial markers' for computer assisted surgery. Fiducial markers are usually titanium screws which are surgically implanted on the bone in the area proximal to the operations and used as 'reference points' by the navigational system of the computer. In our study the titanium screws were replaced by a mix of thermo gelling hydrogel and radio-opaque solutions which were injected in the desired location, without the need for surgical implantation, and then analyzed. Promising results were obtained via computer aided topography.

The research presented in this thesis focuses on the precise molecular design of organic polymers in view of their application in the field of biomaterials.

Nanofibres have potential in biomedical and pharmaceutical applications including tissue engineering and drug release, which demand specific material properties to perform the required function without toxic side-effects, and preferably with minimal adverse ecological impact. Electrospinning is a promising technique for generating fibres with specific requirements and properties from organic, replaceable, non-toxic materials. Aqueous gelatin solution was chosen for its ability to react to changes in temperature. This work, including development of enhanced temperature control, demonstrated nanofibres of gelatin electrospun from aqueous solution with appropriate production conditions, which was previously unreported. To characterise the effects of the gelation mechanism caused by the partial reformation of the triple helix with aqueous gelatin solutions, electrospinning was attempted over a wide range of concentrations and temperatures. The study included measurements of the surface tension, viscosity and conductivity for the solutions, Scanning Electron Microscopy for size and form of product, and Wide Angle X•ray Scattering to ascertain the development of the structures from the solution to the fibre, determining the presence of the triple helix structure, and ThermoGravimetric Analysis to determine fibre water content. To provide comparison and continuity with previous studies, solutions of gelatin dissolved in glacial acetic acid, and polystyrene dissolved in MEK and DMF was electrospun and the products characterised in a similar manner to aqueous gelatin. Polystyrene solutions were chosen as examples of non-gelling solutions, thereby providing a contrast to the aqueous gelatin solutions. A potential area for nanofibres is in drug delivery. Aspirin (acetylsalicylic acid) was considered as an example, but its solubility in water is insufficient to incorporate into electrospun nanofibres for this purpose. Hence, the soluble salt, sodium acetylsalicylate, which shares some of the same medicinal properties as aspirin, was chosen. Nanofibres were electrospun from a solution of gelatin and sodium acetylsalicylate dissolved in water, but then 'disappeared', presumed to have dissolved in air moisture. This demonstrated that gelatin nanofibres could be considered for drug delivery, but further studies would be required to determine methods to stabilise the fibres. Gelatin was dissolved in a suspension of cellulose nanofibres derived from carrots and the resulting liquid was successfully electrospun to produce nanofibres. The nanofibres did not exhibit the expected properties of the gelatin triple helix structure. Attempts to silver coat gelatin nanofibres for medical applications, using Tollens' reagent showed the fibres must be made less soluble by polymer cross-linking or otherwise. A limited study of incorporating gold nanoparticles within the nanofibres to increase their electrical conductivity was inconclusive as the elemental analysis equipment was unable to detect the gold.

Data-driven decision support approaches have been increasingly employed in recent years in order to unveil useful diagnostic and prognostic patterns from data accumulated in clinical repositories. Given the diverse amount of evidence generated through everyday clinical practice and the exponential growth in the number of parameters accumulated in the data, the capability of finding purposeful task-oriented patterns from patient records is crucial for providing effective healthcare delivery. The application of classification decision support tool in clinical settings has brought about formidable challenges that require a robust system. Knowledge Discovery in Database (KDD) provides a viable solution to decipher implicit knowledge in a given context. KDD classification techniques create models of the accumulated data according to induction algorithms. Despite the availability of numerous classification techniques, the accuracy and interpretability of the decision model are fundamental in the decision processes. Multiple Classifier Systems (MCS) based on the aggregation of individual classifiers usually achieve better decision accuracy. The down size of such models is due to their black box nature. Description of the clinical concepts that influence each decision outcome is fundamental in clinical settings. To overcome this deficiency, the use of artificial data is one technique advocated by researchers to extract an interpretable classifier that mimics the MCS. In the clinical context, practical utilisation of the mimetic procedure depends on the appropriateness of the data generation method to reflect the complexities of the evidence domain. A well-defined intelligent data generation method is required to unveil associations and dependency relationships between various entities the evidence domain. This thesis has devised an Interpretable Multiple classifier system (IMC) using the KDD process as the underlying platform. The approach integrates the flexibility of MCS, the robustness of Bayesian network (BN) and the concept of mimetic classifier to build an interpretable classification system. The BN provides a robust and a clinically accepted formalism to generate synthetic data based on encoded joint relationships of the evidence space. The practical applicability of the IMC was evaluated against the conventional approach for inducing an interpretable classifier on nine clinical domain problems. Results of statistical tests substantiated that the IMC model outperforms the direct approach in terms of decision accuracy.

Tissue elasticity assessment is is used in clinical applications from cancer diagnosis to prosthesis fitting. Elastography, using ultrasound to image tissue elasticity to replace digital palpation, includes methods which measure small volume tissue response to high-frequency acoustic excitation and bulk responses following quasi-static loading. Such techniques are well explored and used clinically; however, the goal of quantitative elasticity estimation remains unfulfilled. Additionally, tissue's elastic nonlinearity prevents either approach from providing elasticity information relevant to low-frequency loading as applied by a prosthesis on the lower-limb residuum during gait. This thesis presents research towards the development of a technique utilising the parametric array to exert an acoustic radiation force over small tissue areas under loading frequencies similar to those of gait. The resulting acoustic field magnitude is shown to depend on the elastic properties of the tissue it is developed within. An approach is presented which provides the foundations for a possible medical device to provide high resolution, low frequency, quantitative tissue elasticity information, to aid the prosthetist during fitting or offer the opportunity for automation.

The dialysis membrane is a key component of the haemodialyser. Determination of the membrane solute permeation and ultrafiltration characteristics is important in predicting haemodialyser performance. In some membrane permeation test cells, there are additional mass transfer resistances present due to fluid films adjacent to the membrane surface. This can mask the true permeability of the membrane. Test cell designs should therefore seek to minimise or eliminate any fluid film resistances In this project, solute permeability results were compared using two types of permeation test cells, namely a radial flow cell and a dynamic test cell. In the radial flow cell, fluid streams are directed perpendicular to the membrane surfaces in an attempt to reduce the fluid film thickness. In the dynamic test cell, the fluid streams are passed at high velocity through an open-pore rigid metal support structure that generates localised mixing at the membrane surfaces. Fluid film resistances are proportionately greater for the transfer of low molecular weight solutes and for high flux membranes. Therefore, the transfer of urea (60 mol wt) was studied in a polyacrylonitrile-based membrane AN69ST. Permeability of a higher molecular weight solute, caffeine, (194.2 mol wt) was also determined.