Masters Research - Master of Engineering / This thesis demonstrates the process of building a system to test multiple-input multiple-output (MIMO) communications over-the-air. It covers the entire process, from concept to design and construction, culminating in transmitting space-time coded data packets and producing bit error rate (BER) performance curves. A flexible modular architecture is designed, able to test current MIMO systems and to be upgraded as the field develops. Printed circuit boards for a field-programmable gate array (FPGA) based mainboard, 2.4 GHz transceivers and antennas are then designed, embodying the aforementioned architecture. The mainboard uses a Xilinx XC2S600E FPGA, with ∼600,000 logic gates. Hardware is assembled and tested, forming a foundation for further layers of firmware and software. An abstraction layer, with associated test benches, is written in a hardware description language (VHDL), allowing the core logic of the FPGA to be written and simulated in a device-independent manner. Further VHDL is written and the testbed configured to transmit and receive bursts of data. A device driver is implemented, and abstract data types are layered on top of the driver, enabling high-level control of the testbed. Single antenna and MIMO data links are implemented using 1x1 binary phase-shift keying (BPSK) and 2x2 Alamouti encoded BPSK modulation respectively. Finally, data packets are transmitted and measured BER performance curves constructed. Channel estimation is proved to work on a 2x2 MIMO channel over-the-air, the introduced loss of Eb/N0 shown to be approximately 0.5 dB compared to perfect channel information. The analogue limitations of the hardware are investigated and bit error rate performance measured as a function of operating point. Finally single antenna communications and a 2x2 Alamouti MIMO scheme are compared over-the-air, the Alamouti scheme delivering a 3 dB improvement in Eb/N0 performance. Satisfyingly the MIMO scheme also exceeds the best case theoretical performance bound of the single antenna case by a margin of 2 dB in Eb/N0.

The increased work of breathing associated with intense cycling has been identified as a factor that may negatively affect cycling performance. The aerodynamic position, abnormal respiratory mechanics either at rest or during exercise, and the development of a tachypnoeic breathing pattern are factors known to increase the work of breathing. Breathing pattern retraining aims to decrease the work of breathing by delaying the onset of dynamic hyperinflation and the recruitment of accessory breathing muscles. To date no studies have investigated the performance, physiological and perceptual consequences of manipulating breathing pattern in trained cyclists. Purpose: The aim of the present study was to investigate the effect of breathing pattern retraining on 20-km time trial performance and respiratory and metabolic measures in competitive cyclists. Method: Twenty-four competitive male cyclists (age 37.7 ± 8.6 years, mean ± SD; peak 4.34 ± 0.47 L·min-1) were match paired on 20-km time trial performance and assigned at random to either an intervention group (breathing pattern retraining; N = 12) or control group (N = 12). 20-km time trial performance, pulmonary function and the physiological and perceptual response during a maximal incremental cycle step test were assessed pre- and post-intervention. The intervention group underwent four weeks of specific breathing pattern retraining using exercises designed to reduce dynamic hyperinflation and optimise respiratory mechanics. The control group attended the laboratory once a week during this period and performed a 10 minute sub-maximal ride wearing a biofeedback breathing harness. The control group was led to believe the purpose for their participation was to investigate the effect that maximal exercise had on breathing pattern, and to test the reliability of the breathing harness. There was no attempt to modify the breathing pattern of the control group. Data were analysed using an MS Excel spreadsheet designed for statistical analysis. The uncertainty in the effect was expressed as 90% confidence limits and a smallest worthwhile effect of 1.0% was assumed. Results: The intervention group showed substantial improvements in 20-km time trial performance (-1.5 ± 1.1%) and incremental power (3.2 ± 3%). Additionally, breathing frequency (-13.2 ± 8.9%; -9.5 ± 8.4%), tidal volume (10.6 ± 8.5%; 9.4 ± 7.6%), inspiratory time (10.1 ± 8%; 9.4 ± 7.7%), breathing RPE (-30 ± 33.9%; -24.7 ± 28.1%) and leg RPE (-27.9 ± 38.5%; -24.7 ± 28.2%) were all positively affected at lactate threshold and lactate turn point. No positive changes were observed in the control group for 20-km time trial performance (0.0 ± 1.0%), incremental power (-1.4 ± 3.5%), breathing frequency (-1.6 ± 8.0%; -2.0 ± 7.9%), tidal volume (0.9 ± 7.2%; 2.9 ± 9.4%), breathing RPE (16.1 ± 50.2%, 24.8 ± 43%) or leg RPE (13.4 ± 39.6%; 19.9 ± 43.2%) . Conclusion: These results provide evidence of the performance enhancing effect of four weeks of breathing pattern retraining in cyclists. Furthermore, they suggest breathing pattern can be retrained to exhibit a controlled pattern, without a tachypnoeic shift, during high intensity cycling. Additionally, these results indicate breathing pattern retraining attenuates the respiratory and peripheral perceived effort during incremental exercise. Key words: Breathing pattern disorders, retraining, blood stealing, cycling, performance, power output, respiratory mechanics, perceived exertion, 20km-TT

The incidences of obesity and type 2 diabetes are reaching epidemic proportions worldwide. A cardinal feature of these conditions is resistance to the effects of the hormone insulin and a resulting hepatic overproduction of glucose. Insulin resistance is also implicated in a range of liver diseases including non-alcoholic fatty liver disease (NAFLD) and hepatitis C infection. Insulin is released after a meal and acts on liver, skeletal muscle and adipose tissue to reduce blood glucose concentration. In the liver, insulin inhibits the production and release of glucose into the circulation and stimulates its storage as glycogen. Glucagon, on the other hand, is present in the fasting state and causes breakdown of hepatic glycogen along with production of new glucose. This glucose is released from hepatocytes into the circulation. For the studies in this thesis, functional assays to measure various aspects of hepatic glucose metabolism in vitro were developed. This included measuring glucose output into culture medium, hepatocyte uptake of radiolabelled glucose and incorporation into glycogen, and total cellular glycogen content. These assays were used to investigate glucose metabolism in primary rat hepatocytes and FaO rat hepatoma cells. Both cell types responded to physiological concentrations of insulin, showing decreased glucose output and increased glycogen synthesis. Glucagon increased glucose output and reduced glycogen synthesis in primary cells but had no effect on FaO cells. Factors that have been identified that may inhibit or potentiate insulin action were investigated. Increased body iron stores have been linked with insulin resistance. De-ironing patients improves insulin sensitivity, suggesting a causal relationship between iron and insulin resistance. Hepatocytes store the majority of the body’s excess iron. This project investigated the effects of increasing hepatocyte iron stores, through addition of ferric ammonium citrate (FAC), or depleting iron stores by chelation with dipyridyl. Small increases or decreases of iron in primary cells had negative effects on cell viability, resulting in significantly reduced glucose output and glycogen synthesis. Dipyridyl treatment had similar effects on FaO cells as on primary cells but FAC treatment increased FaO glucose output, although significant iron loading was not achieved. With concentrations of FAC and dipyridyl low enough to not significantly influence cell viability, insulin sensitivity was not affected. Adiponectin is an insulin sensitiser and appears to exert this effect primarily through the liver. Adiponectin can also reduce hepatic glucose output (HGO) independent of insulin. It is believed adiponectin mediates its effects in liver, skeletal muscle and adipose tissue through activation of AMP-activated protein kinase (AMPK). In muscle, p38 mitogen-activated protein kinase (p38 MAPK) has been implicated as a downstream component of adiponectin signalling. In this study, recombinant human adiponectin was produced and collected in culture medium which was then concentrated. Despite the presence of both high molecular weight (HMW) and low molecular weight (LMW) adiponectin multimers, the concentrated medium had no effect on HGO in the presence or absence of insulin. Concentrated adiponectin medium did not affect AMPK or p38 MAPK phosphorylation in hepatocytes or other cell types previously shown to respond to adiponectin. However, commercially-sourced purified recombinant adiponectin also failed to elicit any observable responses. AICAR and metformin are pharmacological activators of AMPK and were used to treat primary rat hepatocytes and FaO cells. These treatments reduced HGO independent of insulin in both cell types. In primary cells, these reductions were partially inhibited with Compound C, an AMPK inhibitor, suggesting that both AICAR and metformin action is at least partly AMPK dependent. In FaO cells, Compound C only inhibited the AICAR-mediated reduction of glucose output, indicating that metformin may act independently of AMPK in these cells. Compound C significantly inhibited AICAR and metformin-mediated increases in AMPK phosphorylation in primary hepatocytes and FaO cells. There was a trend towards inhibition of AICAR-mediated p38 MAPK phosphorylation with Compound C treatment, suggesting that p38 MAPK may lie downstream of AMPK in hepatocytes. Adenoviral expression of constitutively active (CA) and dominant negative (DN) AMPK in primary rat hepatocytes was used to further study the role of AMPK in hepatic glucose metabolism. Despite significant expression of CA AMPK, phosphorylation of downstream acetyl-CoA carboxylase (ACC) was not affected nor was HGO. CA AMPK did, however, increase phosphorylation of p38 MAPK. DN AMPK completely inhibited AICAR-mediated AMPK phosphorylation and partially inhibited phosphorylation of ACC. In addition, AICAR-mediated phosphorylation of p38 MAPK was inhibited by DN AMPK. Taken together, these results suggest that p38 MAPK is downstream of AMPK in hepatocytes. The implication that p38 MAPK is involved in hepatic AMPK signalling is a novel finding. A greater understanding of this pathway in the liver may identify novel therapeutic targets, leading to improved treatment strategies for metabolic disorders linked to obesity and type 2 diabetes.

Drug development is a long and expensive process. It is often not until potential drug candidates are administered to humans that accurate quantification of their pharmacokinetic characteristics is achieved. The goal of developing quantitative structure-pharmacokinetic relationships (QSPkRs) is to relate the molecular structure of a chemical entity with its pharmacokinetic characteristics. In this thesis artificial neural networks (ANNs) were used to construct in silico predictive QSPkRs for various pharmacokinetic parameters using different drug data sets. Drug pharmacokinetic data for all studies were taken from the literature. Information for model construction was extracted from drug molecular structure. Numerous theoretical descriptors were generated from drug structure ranging from simple constitutional and functional group counts to complex 3D quantum chemical numbers. Subsets of descriptors were selected which best modeled the target pharmacokinetic parameter(s). Using manual selective pruning, QSPkRs for physiological clearances, volumes of distribution, and fraction bound to plasma proteins were developed for a series of beta-adrenoceptor antagonists. All optimum ANN models had training and cross-validation correlations close to unity, while testing was performed with an independent set of compounds. In most cases the ANN models developed performed better than other published ANN models for the same drug data set. The ability of ANNs to develop QSPkRs with multiple target outputs was investigated for a series of cephalosporins. Multilayer perceptron ANN models were constructed for prediction of half life, volume of distribution, clearances (whole body and renal), fraction excreted in the urine, and fraction bound to plasma proteins. The optimum model was well able to differentiate compounds in a qualitative manner while quantitative predictions were mostly in agreement with observed literature values. The ability to make simultaneous predictions of important pharmacokinetic properties of a compound made this a valuable model. A radial-basis function ANN was employed to construct a quantitative structure-bioavailability relationship for a large, structurally diverse series of compounds. The optimum model contained descriptors encoding constitutional through to conformation dependent solubility characteristics. Prediction of bioavailability for the independent testing set were generally close to observed values. Furthermore, the optimum model provided a good qualitative tool for differentiating between drugs with either low or high experimental bioavailability. QSPkR models constructed with ANNs were compared with multilinear regression models. ANN models were shown to be more effective at selecting a suitable subset of descriptors to model a given pharmacokinetic parameter. They also gave more accurate predictions than multilinear regression equations. This thesis presents work which supports the use of ANNs in pharmacokinetic modeling. Successful QSPkRs were constructed using different combinations of theoretically-derived descriptors and model optimisation techniques. The results demonstrate that ANNs provide a valuable modeling tool that may be useful in drug discovery and development.

Thesis (Ph. D.)--University of Washington, 2007. / Vita. Includes bibliographical references (p. 127-131).

Thesis(M.S.)--Auburn University, 2006. / Abstract. Vita. Includes bibliographic references.

Thesis (PhD (Philosophy))-- University of Western Sydney, Nepean, 1998. / "Thesis submitted for the fulfilment of the requirement of Doctor of Philosophy in quantitative methods, School of Quantitative Methods and Business Operations, Faculty of Commerce, University of Western Sydney, Nepean" Bibliography : p 233-257.

Thesis (M.S.)--Missouri State University, 2008. / "May 2008." Includes bibliographical references (leaves 22-23). Also available online.

Thesis (M.Sc. (Physiology))--Mahidol University, 1978.

Lic.-avh. (sammanfattning) Stockholm : Karolinska institutet, 2007. / Härtill 2 uppsatser.