Wearable sensor systems composed of small and light sensing nodes have the potential to revolutionise healthcare. While uptake has increased over time in a variety of application areas, it has been slowed by problems such as lack of infrastructure and the functional capabilities of the systems themselves. An important application of wearable sensors is the detection of falls, particularly for elderly or otherwise vulnerable people. However, existing solutions do not provide the detection accuracy required for the technology to gain the trust of medical professionals. This thesis aims to improve the state of the art in automated human fall detection algorithms through the use of a machine learning based algorithm combined with novel data annotation and feature extraction methods. Most wearable fall detection algorithms are based on thresholds set by observational analysis for various fall types. However, such algorithms do not generalise well for unseen datasets. This has thus led to many fall detection systems with claims of high performance but with high rates of False Positive and False Negative when evaluated on unseen datasets. A more appropriate approach, as proposed in this thesis, is a machine learning based algorithm for fall detection. The work in this thesis uses a C4.5 Decision Tree algorithm and computes input features based on three fall stages: pre-impact, impact and post-impact. By computing features based on these three fall stages, the fall detection algorithm can learn patterns unique to falls. In total, thirteen features were selected across the three fall stages out of an original set of twenty-eight features. Further to the identification of fall stages and selection of appropriate features, an annotation technique named micro-annotation is proposed that resolves annotation-related ambiguities in the evaluation of fall detection algorithms. Further analysis on factors that can impact the performance of a machine learning based algorithm were investigated. The analysis defines a design space which serves as a guideline for a machine learning based fall detection algorithm. The factors investigated include sampling frequency, the number of subjects used for training, and sensor location. The optimal values were found to be10Hz, 10 training subjects, and a single sensor mounted on the chest. Protocols for falls and Activities of Daily Living (ADL) were designed such that the developed algorithms are able to cope under a variety of real world activities and events. A total of 50 subjects were recruited to participate in the data gathering exercise. Four common types of falls in the sagittal and coronal planes were simulated by the volunteers; and falls in the sagittal plane were additionally induced by applying a lateral force to blindfolded volunteers. The algorithm was evaluated based on leave one subject out cross validation in order to determine its ability to generalise to unseen subjects. The current state of the art in the literature shows fall detectors with an F-measure below 90%. The commercial Tynetec fall detector provided an F-measure of only 50% when evaluated here. Overall, the fall detection algorithm using the proposed micro-annotation technique and fall stage features provides an F-measure of 93% at 10Hz, exceeding the performance provided by the current state of the art.

Medical diagnosis is the process of attempting to determine and/or identify a possible disease or disorder. This process is revealed by biomarkers, defined by The Food and Drug Administration (FDA) as “characteristics that are objectively measured and evaluated as indicators of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention”. The process of biomarker discovery has been boosted in the last years by proteomics, a research discipline that takes a snapshot of the entire wealth of proteins in an organism/ tissue/ cell/ body fluid. An implementation of the analysis methods can help in isolate proteins present in the low range of concentrations, such as biomarkers very often are. An established biomarker can further be measured with the help of biosensors, devices that can be employed in the point-of care diagnostics. This PhD thesis shows and discusses the results of three projects in the field of protein biomarkers discovery and quantification. The first project exploited proteomics techniques to find relevant protein markers for Intrauterine Growth Restriction (IUGR) in cordonal blood serum (UCS) and amniotic fluid (AF). A 14 proteins in UCS and 11 in AF were successfully identified and found to be differentially expressed. Molecularly Imprinted Polymers (MIPs) directed towards proteins and peptides containing phosphotyrosine were then produced, with the final goal of selectively extracting phosphopeptides from a peptide mixture. An alteration of the phosphorylation pattern is in fact often associated to important diseases such as cancer. The polymers were produced as nanoparticles, that were characterised with Dynamic Light Scattering (DLS) and Atomic Force Microscopy (AFM). A recipe was also tested for binding capacity towards phosphotyrosine. A Surface Plasmon Resonance (SPR) biosensor to quantify hepcidin hormone was finally produced. This is the major subject in iron homeostasis in vertebrates and marker of iron unbalance diseases. A calibration curve was made and affinity/kinetic parameters for the ligand employed were measured.

Interactions between single and double stranded oligonucleotides with SiO2 surfaces and the interactions between oligonucleotides and immobilized oligomers have been studied computationally. The oligonucleotide is the 18-base-pair sequence for the survival motor neuron gene SMN1. The oligomer consisted of a 50 unit 2-hydroxyethyl methacrylate (PHEMA) molecule. A linker used to tether the oligonucleotide was either a 10 Å or a 30 Å long succinimdyl 4-[N-maleimidomethyl]cyclohexane-1-caroxylate (sulfo-SMCC-Cn). The surface consisted of a SiO2 crystal that was 50 Å long and 50 Å wide, one unit thick and covered with modified-(3-aminopropyl)trimethoxysilane (m-APTMS) molecules. It was determined that explicit water, sodium counterions and excess salt were necessary to produce computationally stable oligonucleotide structures on surfaces. Artificial partial charges were introduced to the surface, and linkers, oligomers and oligonucleotides were immobilized and studied. The linkers collapsed onto a positive but not onto a negative surface. Oligomers moved closer to the SiO2 surface regardless of the surface charge. Immobilized oligonucleotides tilted significantly from an initial upright position but did not collapse completely onto the surfaces. The interactions between immobilized oligonucleotides and oligomers were examined. The number of oligomers surrounding the oligonucleotide was varied between two and four. Single stranded oligonucleotides were prevented from interacting with the surface as they were inhibited by the presence of oligomers. Double stranded oligonucleotides collapsed onto the surface when only two oligomers were present but remained upright when four oligomers were present. This was due to the four oligomers interacting with one another and effectively shielding the surface. The oligomers interacted with the bases in the single stranded oligonucleotides, making them energetically accessible. Presence of a high density of oligomers prevented the dsDNA from collapsing onto the surface. These results suggest design criteria for preparation of mixed oligonucleotide and oligomer films for use in biosensors.

Adverse interaction and subsequent adsorption of biomolecular species (i.e. fouling) pose a great hindrance for medical and clinical applications (e.g. biosensors). Research into the mechanism behind antifouling coatings have shown a strong link between surface hydration and antifouling behaviour due to the existence of a ‘water barrier’ which prevents proteins from adsorbing onto the surface. In a previous study, a short, mono(ethylene-glycol) silane adlayer (MEG-OH) showed significantly different antifouling behaviour in comparison to its homolog – lacking the internal ether oxygen (OTS-OH). In the present work, neutron reflectometry (and modeling) was used to investigate the water density profiles at MEG-OH and OTS-OH silane adlayers on quartz and Si/SiO2 to determine whether the internal ether oxygen affects the adlayers’ interaction with water. Despite the limitations of studying such ultrathin organic films, the two systems showed different hydration profiles supporting the link between surface hydration and antifouling.

The aim of this study was to develop a suitable monitoring protocol for mediated amperometric whole-cell biosensors for in situ assessment of heavy metals in soil. E. coli 8277, Pseudomonas 9773, Pseudomonas 9046 and Pseudomonas 8917 were screened as biosensor catalysts to select the sensitive biosensor configurations to heavy metals. A new protocol was developed for monitoring heavy metals in defined solution, soil pore water, and in situ in soil. This study also demonstrated the applications of mediated amperometric bacterial biosensors for in situ assessing the bioavailability and toxicity of heavy metals in freshly spiked soils or historically contaminated soils, and mixture toxicities of heavy metals. It was found that the biosensors incorporating selected bacterial strains were appropriately sensitive to copper, but less sensitive to Zn, Pb, and Hg, compared to Microtox assay. The advantage of the mediated amperometric bacterial biosensor system is its in situ application in soils. The present study demonstrated that soil pore water does not accurately reflect conditions of soil ecosystem, and that in situ bioassays are more reliable for determining the bioavailability and toxicity of heavy metals. This is the first reported use of disposable whole cell biosensors for in situ heavy metal bioavailability and toxicity assessment. The biosensor protocol developed here can be adapted to allow the incorporation of dfferent bacterial biocatalysts for applications in soil quality assessment, screening of sites for contamination ‘hot spots’, and the evaluation of soil degradation or rehabilitation from metal pollution. Mediated amperometric bacterial biosensors are not analyte specific, their response reflecting the metabolic impact of the combined chemical and physical properties of the environment to which they are exposed. In assessing the toxicity of soil samples from fields using these biosensors, it is vital to get appropriate control soil samples. The conditions of soil samples also need to be well defined. The sensitivity of the mediated amperometric whole-cell biosensors to heavy metals need to be further improved. Investigations are also required to determine how the natural conditions affect the application of the biosensor system in the field.

The performance of a novel, visible light-driven photoelectrocatalytic (PEC) batch reactor employing tungsten trioxide (WO₃) as a photocatalyst was assessed by studying the degradation of selected model pollutants (2,4-DCP, chloroform) and the disinfection of a human bacterial pathogen (<i>E. coli </i>O157:H7). Overall efficacy of the batch reactor was assessed by combining biological toxicity assessment (biosensing) with conventional analytical chemistry. Photoelectrocatalytic degradation of the organoxenobiotics (2,4-DCP, chloroform) was monitored toxicologically by applying bacterial <i>lux</i>-marked biosensors and analytically by HPLC. The bacterial biosensor traced the removal of the target, model pollutants during degradation experiments, and also monitored changes in toxicity in the analyte of the PEC batch reactor caused by the possible appearance/disappearance of toxic transient intermediates derived from the breakdown of the parent molecule. Chromosomally <i>lux</i>-marked, non-toxigenic <i>E. coli</i> O157:H7 was selected as a model human pathogenic bacterium to demonstrate the disinfection potential of the batch reactor. Results of disinfection experiments indicated that a substantial decline in the population density of culturable <i>E. coli </i> O157:H7 cells was achieved. Accurate differentiation between the effects of photoelectrocatalysis and photolysis on the cells of <i>E. coli</i> O157:H7 was not achieved. The observed rate of the degradation of the model chemical compounds and the disinfection of the model human pathogen, demonstrated that visible light-driven photoelectrocatalysis offers considerable potential for remediation of contaminated water. Furthermore, toxicological biosensing can bridge the gap between traditional chemical analysis and ecologically relevant sample evaluation and address suitability of reintroduction of treated solution back into mainstream wastewater treatment.

Submitted in fulfillment of the requirements for the Degree of Master of Applied Science in Chemistry, Durban University of Technology, Durban, South Africa, 2016. / This study involves the development of a sensitive electrochemical biosensor for the determination of capsaicin extracted from chilli pepper fruit, based on a novel signal amplification strategy. The study therefore, seeks to provide a sensitive electro-analytical technique to be used for the determination of capsaicin in food and spicy products. Electrochemical measurements using cyclic voltammetry (CV) and differential pulse voltammetry (DPV) modes were utilized in order to understand the redox mechanism of capsaicin and to test the performance of the developed biosensor supported with computational techniques. In this work two different enzymes, Phenylalanine ammonia lyase (PAL) and Glucose oxidase (GOx) were used for electrode modifications respectively. For this purpose three different types of working electrodes namely: glassy carbon electrode (GCE), platinum electrode (Pt-E) and gold electrode (Au-E) were used and their performances were compared. For the first time, the three electrodes were modified with PAL and GOx enzymes on multiwalled carbon nanotubes used in this study and characterized by attenuated total reflectance infrared spectroscopy, transmittance electron microscopy and thermo-gravimetric analysis supported by computational methods. The comparison of the results obtained from the bare and modified platinum electrodes revealed the sensitivity of the developed biosensor with modified electrode having high sensitivity of 0.1863 µg.L-1 and electron transfer rate constant (ks) of 3.02 s-1. To understand the redox mechanism completely, adsorption and ligand-enzyme docking simulations were carried out. Docking studies revealed that capsaicin formed hydrogen bonds with Glutamates (GLU355, GLU541, GLU586), Arginine (ARG) and other amino acids of the hydrophobic channel of the binding sites which facilitated the redox reaction for detection of capsaicin. These results confirm that the PAL enzyme facilitated the electron transfer from the capsaicin ligand, hence improving the biosensing response. Our results suggest potential applications of this methodology for the determination of capsaicin in the food industry. / M

M.Tech. (Chemistry) / Please read abstract in the full-text document

Orientador: Paulo Roberto Bueno / Coorientador: Maria Del Pilar Taboada Sotomayor / Banca: Hideko Yamanaka / Banca: Emanuel Carrilho / Banca: Marcelo Mulato / Banca Mauro Bertotti / Resumo: A técnica de espectroscopia de capacitância eletroquímica foi recentemente utilizada com sucesso para detecção de biomarcadores de interesse clínico, como o caso da proteína C-reativa (CRP), que está relacionada com doenças cardíacas e processos inflamatórios. Nesta tese, esta técnica foi utilizada para desenvolver dispositivos eletroanalíticos com possíveis aplicações para a detecção de trombose, utilizando a proteína biomarcadora D-dímero; câncer de próstata, por meio da detecção da enzima fosfatase ácida prostática (PAP), e glicoproteína HRP, para possível detecção de células tumorais e como modelo para o estudo da interação lectina-glicoproteína. Além, no último caso, utilizaram-se também as funções de imitância como sinal transdutor como possível aplicação em glycoarray. Para a construção da superfície receptora sobre eletrodo de ouro, foram utilizadas duas moléculas diferentes de tiol (R-SH, sendo R uma cadeia carbônica genérica). A primeira, que contém em sua estrutura o grupo terminal carboxílico, foi utilizada com o objetivo de imobilizar o material biológico para reconhecimento do analito (i.e., anticorpos ou lectina) via protocolo EDC/NHS. A segunda, com o grupo redox terminal (11-ferrocenil-undecanotiol, 11Fc), foi utilizada com o intuito de fornecer o sinal transdutor. As monocamadas bifuncionais obtidas pela adsorção conjunta de ambas as moléculas apresentam densidade molecular superficial na ordem de 5 x 10-10 mol/cm2. A imobilização do anticorp... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The electrochemical capacitance spectroscopy technique has been recently applied to detect biomarkers of clinical interest, such as C-reactive protein (CRP) which is related to heart disease and inflammatory processes. Herein, this technique was used to develop electroanalytical approaches with potential applications for thrombosis diagnosis by detecting D-dimer biomarker protein; prostatic cancer, using prostatic acid phosphatase (PAP) enzyme as target; and glycoprotein assay, for possible detection of tumor cells and as a model for studying lectin-glycoprotein interaction (using ArtinM lectin and HRP glycoprotein). In addition, in the latter case, immittance functions were used as transducer signals as possible application for glycoarray. For the construction of the receptor surface on gold electrodes, two different thiol molecules (R-SH, being R a generic carbonic chain) were used. The first, which contains the carboxylic terminal group, was applied to immobilize the biological material for target recognition (i.e., either antibodies or lectin) via EDC/NHS protocol. The second molecule contains a terminal redox group (11-ferrocenyl-undecanethiol, 11Fc) in order to provide the transducer signal. The bifunctional monolayers obtained herein present high surface coverage, ≈ 5 x 10-10 mol/cm2. The immobilization of the antibody to PAP detection (anti-PAP) was investigated by QCM, and it was verified that saturation occurs in approximately 1 h, yielding a surface coverage of ≈ 3.7 mg/m2, suggesting "end on" orientation. Using the electrochemical capacitance signal, it was possible to develop an approach for PAP detection with limit of detection (LD) and quantification (LQ) of 9 pmol/L and 28 pmol/L, respectively, applied to a PAP clinically useful concentration range of 50-1000 pmol/L in phosphate buffered saline (PBS, pH 7.4). The relative standard deviation obtained was 9.5%, and showed specif... / Doutor

Orientador: Hideko Yamanaka / Co-orientador: Antonio Aparecido Pupim Ferreira / Banca: Assis Vicente Benedetti / Banca: Zeki Naal / Resumo: Os corantes azo estão entre os mais utilizados pela indústria têxtil brasileira, representando de 20-40% dos corantes empregados para tingir algodão, rayon, nylon, seda, lã e couro. Um fator preocupante relacionado a estas substâncias é que estudos utilizando microorganismos e células de mamíferos têm demonstrado que diversos corantes azo apresentam atividade genotóxica, mesmo em baixas concentrações. Além disso, atualmente ainda não estão vigorando métodos oficiais para efetiva remoção dos corantes presentes nos efluentes gerados no processo de tingimento dos tecidos, e isso faz com que essas substâncias possam chegar à agua destinada ao consumo, e assim, o desenvolvimento de dispositivos sensores para o monitoramento dos corantes em água se torna de grande importância. Diante disso, o presente trabalho versa sobre o desenvolvimento de um imunossensor impedimétrico para detecção e quantificação do corante azo Disperso Red 1 (DR1) em baixos níveis de concentração em água tratada e para tal, duas estratégias metodológicas foram estudadas. A primeira delas se fundamentou na construção do imunossensor empregando eletrodos impressos de ouro modificados com monocamadas auto-organizadas. Monocamadas tioladas de cistamina, ácido lipóico e p-aminotiofenol foram estudadas para promoção do acoplamento dos anticorpos sobre a superfície do eletrodo. Caracterizações realizadas por voltametria cíclica (VC) e espectroscopia de impedância eletroquímica (EIS) demonstraram a efetiva imobilização dos anticorpos sobre a superfície dos eletrodos impressos, com aumento sucessivo do recobrimento e no valor de resistência de transferência de carga (Rct), respectivamente. No entanto, para todas as rotas estudadas encontraram-se dificuldades em se obter adequada repetibilidade entre as diferentes medidas e estabilidade das etapas de modificação, além de não ser observada... / Abstract: Azo dyes are among the most used by Brazilian textile industry, representing 20-40% of dyes used for dyeing cotton, rayon, nylon, silk, wool and leather. A worrying factor related to these substances is that studies using micro-organisms and mammalian cells have shown that many azo dyes have genotoxic activity, even at low concentrations. Besides, currently there are not oficial methods for effective removal of dyes present in the effluents generated in the dyeing process, and this leads these substances can reach the water for consumption, and thus the development of sensor devices for monitoring the dyes in water becomes of great importance. Therefore, this project is focused on development of an impedimetric immunosensor for detection and quantification of the azo dye Disperse Red 1 (DR1) at low concentration levels in treated water and for this, two methodological strategies were studied. The first one was based on the construction of the immunosensor employing gold printed electrodes modified with self-assembled monolayers. Thiolated monolayers of cystamine, lipoic acid and p-aminothiophenol were studied to promote the coupling of the antibodies on the electrode surface. Characterizations performed by cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) demonstrated the effective immobilization of antibodies on the surface of the printed electrodes, with successive increase in the surface coverage and in the charge transference resistance (Rct), respectively. However for all ways investigated, difficulties in obtaining adequate repeatability between measures of stability and modification steps were observed, as well a non-linear relationship between the values of ΔRct and the concentration of antigen DR1, which was due by the heterogeneity of the surface of different printed electrodes. Thus, the second strategy employed glassy carbon electrode for immunosensor construction... / Mestre