Spectroscopic analysis has several applications in physics, chemistry, bioinformatics, geophysics, astronomy, etc. It has been widely used for detecting mineral samples, gas emission, and food volatiles. Machine learning algorithms for spectroscopic analysis focus on either regression or single-label classification problems. Using multi-label classification to identify multiple chemical components from the spectrum, has not been explored. In this thesis, we implement Feed-forward Neural Network with Optimal Thresholding (FNN-OT) identifying gas species among a multi gas mixture in a cluttered environment. Spectrum signals are initially processed by a feed-forward neural network (FNN) model, which produces individual prediction scores for each gas. These scores will be the input of a following optimal thresholding (OT) system. Predictions of each gas component in one testing sample will be made by comparing its output score from FNN against a threshold from the OT system. If its output score is larger than the threshold, the prediction is 1 and 0 otherwise, representing the existence/non-existence of that gas component in the spectrum. Using infrared absorption spectroscopy and tested on synthesized spectral datasets, our approach outperforms FNN itself and conventional binary relevance - Partial Least Squares with Binary Relevance (PLS-BR). All three models are trained and tested on 18 synthesized datasets with 6 levels of \signal-to-noise ratio and 3 types of gas correlation. They are evaluated and compared with micro, macro and sample averaged precision, recall and F1 score. For mutually independent and randomly correlated gas data, FNN-OT yields better performance than FNN itself or the conventional PLS-BR, by significantly by increasing recall without sacrificing much precision. For positively correlated gas data, FNN-OT performs better in capturing information of positive label correlation from noisy datasets than the other two models. / Graduate

La spectrométrie de masse tandem est une technique analytique versatile, notamment utilisée dans le champ de la protéomique pour dériver la séquence de peptides. Cette thèse vise à contribuer au développement de méthodes intégrées à la spectrométrie de masse afin d’apporter des informations structurales sur les ions moléculaires, intermédiaires réactionnels ou produits de réactions. La fragmentation des peptides protonés est induite par collisions avec une gaz rare (CID). Des multiples fragments sont observés et la séquence peptidique peut être dérivée de la mesure de la masse de séries d’ions analogues. Les mécanismes de fragmentation par CID des peptides protonés sont très complexes, constituant un champ de recherche important. Dans ce contexte, le couplage de deux techniques comme la spectrométrie de masse et la spectroscopie infrarouge a été utilisé pour caractériser des ions fragments (an et bn) de peptides. Divers signatures infrarouges identifiées dans les gammes spectrales 1000-2000 et 3000-3800 cm-1 et caractéristiques de différents motifs structuraux ont mis en évidence la permutation de la séquence de la chaîne peptidique pour les ions an. Dans le cas des ions bn, la formation d’une structure macrocyclique s’appuie sur la formation spécifique d’un complexe produit d’une réaction avec NH3, que nous avons caractérisé par spectroscopie infrarouge. L’objectif ultime de cette thèse était de mettre en place une approche multimodale sur un spectromètre de masse unique, intégrant la séparation par mobilité ionique et la spectroscopie infrarouge d’ions moléculaires en permettant la caractérisation structurale par spectroscopie infrarouge des ions simultanément sélectionnés en masse et par mobilité ionique. La technique de mobilité ionique mise en œuvre est de type « Differential Ion Mobility spectrometry » (DIMS). Nous aurions souhaité pouvoir exploiter l’IMS sur des ions fragments de peptides, mais la technique DIMS ne le permettait pas. Nous avons choisi d’explorer la séparation et la caractérisation des monosaccharides cationisés par Li+, Na+ et K+. Dans le cas des complexes de Li+, les résultats spectroscopiques et de mobilité ionique sont cohérents avec les structures les plus stables prédites par la théorie. Ces calculs de chimie quantique permettent aussi d’interpréter une signature spectrale spécifique de complexes de Li+ avec des anomères de glucose. / Tandem mass spectrometry is a versatile analytical technique, used in particular in the field of proteomics to derive peptide sequence. This thesis aims to contribute to the development of integrated approaches to mass spectrometry to provide structural information on molecular ions, which can either be reaction intermediates or reaction products. The fragmentation of protonated peptides is induced by multiple collisions with rare gas atoms (CID). Multiple fragments are observed and the peptide sequence is derived from the measurement of the mass difference between two consecutive analogous ions. Fragmentation mechanisms of the protonated peptides under CID conditions constitute an intense research field. In the frame of this PhD thesis, mass spectrometry and infrared spectroscopy were coupled to characterize the fragment ions (an and bn) peptides. Several infrared signatures were found in the 1000-2000 and 3000-3800 cm-1 spectral ranges characteristic of different structural motifs. In the case of the an ions, these IR signatures provide evidence of the permutation of the sequence of the peptide chain. In the case of the middle size bn ions, the formation of a macrocyclic structure relies on the specific formation of an ion-molecule complex with NH3, whose structure has been characterized by IR spectroscopy.The ultimate objective of this thesis was to develop a multimodal approach based on a single mass spectrometer, which incorporates ion mobility spectrometry and infrared spectroscopy in order to characterize the structure of mobility- and mass-selected molecular ions. The ion mobility technique used is the "Differential Ion Mobility spectrometry" (DIMS). Peptide fragment ions could not be studied using this new set-up. We chose to study the separation and characterization of cationized monosaccharides with Li+, Na+ or K+. In the case of lithiated complexes, the spectroscopic and ion mobility data are consistent with the low energy structures predicted by theory, allowing in particular the interpretation of an IR specific signature characteristic of some Li+ complexes of glucose anomers.

Orientador: Sandra Helena Unêda Trevisoli / Coorientador: Antônio Orlando Di Mauro / Coorientador: Danilo Luiz Flumignan / Banca: Luciana Maria Saran / Banca: Everton Luis Finotto / Resumo: Objetivou-se com este trabalho construir uma curva de calibração para predição do teor de óleo de pinhão-manso, usando a espectroscopia no infravermelho próximo (NIR), bem como traçar o perfil composicional do óleo de pinhão-manso por cromatografia gasosa, além de utilizar técnicas multivariadas para seleção de acessos promissores para programas de melhoramento genético. Para construção da curva, foram utilizadas sementes de 142 plantas, oriundas de 56 acessos, que tiveram óleo extraído por método destrutivo de sementes, para obtenção dos valores de referência, que situaram-se entre 29,32 a 49,69% (m/m) de óleo. Os espectros foram coletados de sementes sem tegumento, intactas em equipamento de espectroscopia NIR. Cada espectro foi obtido com uma média de 64 scans com resolução de 16 cm-1. As regiões espectrais selecionadas para a curva de calibração foram de 8344 a 7736 cm-1, 6072 a 4968cm-1 e 4520 a 3976cm-1. Para avaliar a precisão da curva foram calculados o coeficiente de correlação (R2) que foi de 71,26% e o erro padrão de calibração (RMSECV) que foi de 0,138 entre os dados obtidos pelo método de referência e pelo NIR. Os resultados obtidos indicam que o método de referência utilizado foi satisfatório, apresentando boa correlação com a informação espectral, tornando a curva construída confiável, necessitando, entretanto da adição de novos dados para incremento na sua robustez e amplitude. A cromatografia gasosa indicou que os ácidos graxos com maiores porcentagens foram: oléico (43,45%), linoléico (35,13%), palmítico (13,44%) e esteárico (6,30%), concluindo que o perfil composicional do pinhão-manso é predominantemente insaturado. Para a abordagem multivariada, foram obtidos dados dos seguintes caracteres: teor de óleo, diâmetro do caule, altura da planta, comprimento e largura de semente, sendo os mesmos submetidos a análise de ... / Abstract: The aim of this study was to establish a calibration curve for prediction of oil content by NIR and the compositional profile by gas chromatography, both of Jatropha curcas, such as analyze multivariate techniques for purposes of selection of promising accessions for breeding programs. For the development of calibration curve, were used seeds of 142 plants, originated of 56 accessions, which had oil extracted by destructive method of seed to obtain reference values, which ranged from 29.32 to 49.69% of oil. The spectra were collected from intact seed's kernels. Each spectrum was taken with an average of 64 scans with resolution of 16 cm-1.The regions established for the calibration were 8344-7736 cm-1, 6072-4968 cm-1 and 4520-3976 cm- 1. To evaluate the accuracy of the curve, was calculated the correlation coefficient (R2) which was 71.26% and the standard error of calibration (RMSECV) which was 0.138 between the data obtained by the reference method and the NIR. The results indicate that the reference method was satisfactory, showing correlation with the spectral information, making the curve reliable, requiring however, the addition of new data to increase the robustness and extent. Gas chromatography showed that fatty acids with the highest percentages were: oleic (43.45%), linoleic (35.13%), palmitic (13.44%) and stearic (6.30%), concluding that the compositional profile of Jatropha curcas oil is predominantly unsaturated. For multivariate approach data were obtained from the traits: oil content, stem diameter, plant height, length and width of seed. The data were submitted to principal component analysis and cluster analysis by the hierarchical method using Euclidean distance and connection between groups by Ward method. In principal components analysis, three components explained 74.1% of the variance contained in the original variables and were characterized by the variable oil ... / Mestre

Orientador: Sandra Helena Unêda-Trevisoli / Coorientador: Antônio Orlando Di Mauro / Banca: Gustavo Henrique de Almeida Teixeira / Banca: Ivana Marino Bárbaro / Resumo: O objetivo do presente trabalho consistiu em caracterizar a diversidade fenotípica em genótipos de pinhão-manso através de análises multivariadas, além de desenvolver um modelo de calibração para a determinação do teor de óleo por meio da espectroscopia do infravermelho próximo (NIR) visando a seleção de genótipos com alto teor de óleo. Para a análise fenotípica, foram avaliados 11 caracteres agronômicos em 165 genótipos de pinhão-manso pertencentes ao banco de germoplasma in vivo localizado no Departamento de Produção Vegetal da UNESP/FCAV, Jaboticabal, SP. Os caracteres utilizados foram: altura da planta, diâmetro do caule, número de ramificações primárias, comprimento de fruto, largura de fruto, peso de fruto, formato de fruto, comprimento de semente, largura de semente, peso de semente e teor de óleo. Os dados foram submetidos à análise de componentes principais (PCA), análise de agrupamento pelo método hierárquico de Ward e método não hierárquico k-médias. O caráter formato de fruto foi retirado das análises multivariadas por ser de natureza qualitativa. Na PCA, os 4 CPs explicaram 71,62% da variância total. Os caracteres de maior contribuição para a discriminação dos genótipos no CP1 foram peso de semente, largura de fruto, comprimento de fruto e peso de fruto. Destacaram-se 22 genótipos promissores com potencial para serem explorados em programas de melhoramento genético. O dendrograma obtido pela análise de agrupamento de Ward e para a análise de k-médias geraram nove... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The purpose of this study was to characterize the phenotypic diversity in Jatropha genotypes by multivariate analysis, and developing a calibration model for determination oil content by Near-infrared spectroscopy (NIR) in order to select genotypes with high oil content. For the phenotypic analysis, 11 agronomic characters were evaluated in 165 genotypes of Jatropha belonging to the germplasm bank in vivo located in the Department of Plant Production UNESP / FCAV, Jaboticabal, SP. The characters were: plant height, stem diameter, number of primary branches, fruit length, fruit width, fruit weight, fruit shape, seed length, seed width, seed weight and oil content. The data were submitted to principal component analysis (PCA), cluster analysis by the hierarchical method of Ward and non- hierarchical method of k-means. The character fruit shape was removed from multivariate analysis for being qualitative. In PCA, the 4 PCs explained 71.62% of the total variance. The greatest contribution of the characters to discriminate the genotypes in PC1 were seed weight, fruit width, fruit length and fruit weight. They highlighted 22 promising genotypes with potential to be exploited in breeding programs. The dendrogram obtained by Ward cluster analysis and the k-means analysis generated nine groups, 5 of which grouped genotypes with advantageous features. Multivariate analyzes allowed the characterization of the genotypes, allowing the selection for many traits and indicating gains strateg... (Complete abstract click electronic access below) / Mestre

Os catalisadores eletroquímicos binários de PtSn/C, PdSn/C e PtPd/C foram sintetizados em diferentes proporções pelo método da redução via borohidreto, posteriormente estes foram caracterizados por microscopia eletrônica de transmissão, difração de raios X, espectroscopia no infravermelho por transformada de Fourier (PtSn/C e PdSn/C) e energia dispersiva de raios X. As atividades eletroquímicas dos diferentes materiais preparados foram avaliadas por intermédio de voltametria cíclica, cronoamperometria e curvas de polarização em célula a combustível alimentada diretamente por etileno glicol em eletrólito alcalino. As curvas de densidade de potência indicaram que os catalisadores eletroquímicos contendo Sn e Pd são mais ativos para a reação de oxidação do etileno glicol, especialmente a composição 70%:30% - relação molar entre os metais suportados em carbono - dos catalisadores PtSn/C, PdSn/C e PtPd/C todos superando as medidas de potência do Pt/C. Este resultado indica que a adição de Sn e Pd favorece a oxidação do etileno glicol em meio alcalino. O melhor desempenho observado para os catalisadores eletroquímicos PtSn/C, PdSn/C e PtPd/C (70%:30%) poderia estar associado à sua maior seletividade quanto a formação de oxalato, ou seja , a formação deste produto resulta em um maior número de elétrons, por consequência em maiores valores de corrente. / Binary electrochemical catalysts PtSn/C, PdSn/C and PtPd/C were synthesized in different proportions by the method of reduction via borohydride. These were characterized by transmission electron microscopy, X-ray diffraction, Fourier transform infrared spectroscopy (PtSn/C and PdSn/C) and energy dispersive X-ray. The electrochemical activities of the different materials prepared were evaluated by cyclic voltammetry, chronoamperometry and polarization curves for fuel cell powered directly by ethylene glycol in an alkaline electrolyte. Power density curves indicated that the electrochemical catalysts Sn-containing or Pd-containing are more active for ethylene glycol oxidation reaction, particularly the (70%:30%) composition of PtSn/C, PdSn/C and PtPd/C, all of them exceeding power measurements of Pt/C. These results indicate that the addition of Sn and Pd promotes the oxidation of ethylene glycol in an alkaline medium and this improved performance may be associated with a higher selectivity for the formation of oxalate that results in a larger number of electrons, consequently enhancing the current values.

A adição de corantes fluorescentes a adesivos odontológicos possibilita a investigação da distribuição espacial desses materiais na interface dente-restauração, utilizando-se a microscopia confocal de varredura a laser (MCVL). A literatura indica falta de padronização na aplicação de agentes fluorescentes com tal finalidade. Esse estudo sistematizou estratégias para a adição de rodamina B (RB) e fluoresceína sódica (FS) a um sistema adesivo convencional de três passos, Adper Scotchbond Multi-Purpose (MP), e um autocondicionante de dois passos, Clearfil SE Bond (SE), considerados padrão-ouro na Odontologia. Os objetivos principais foram (a) determinar a menor faixa de concentrações de RB e FS necessária para produzir imagens satisfatórias da interface dentina-adesivo e (b) avaliar o efeito da adição desses corantes sobre algumas propriedades das resinas. Os adesivos foram marcados com RB ou FS em concentrações decrescentes (0,5, 0,1, 0,02 e 0,004 mg/mL) por meio de um método de dispersão semidireto. O comportamento fotofísico/ fluorescente dos adesivos marcados foi investigado por espectroscopia de fotoluminescência e MCVL. Paralelamente, avaliaram-se os adesivos quanto ao grau de conversão (GC) e ao ângulo de contato (AC). Tanto os resultados de GC como os de AC foram submetidos à análise de variância com dois fatores (adesivo e tratamento) com = 0,05, seguida de teste post-hoc de Tukey. Os máximos comprimentos de onda de emissão e de excitação da RB e da FS foram influenciados pelo meio polimérico e pela concentração de corante de modo geral. A MCVL preliminar de amostras de adesivo polimerizado, realizada sob condições experimentais padronizadas, mostrou que o comportamento fluorescente da RB em MP e SE foi muito semelhante na mesma concentração de corante, mas o mesmo não pôde ser dito do comportamento da FS, que foi notavelmente inferior no adesivo autocondicionante, SE, na concentração mais alta. Em dentina, os adesivos preparados com RB nas concentrações-alvo de 0,1 e 0,02 mg/mL apresentaram fluorescência ótima; já aqueles preparados com 0,004 mg/mL produziram fraco sinal. Adesivos preparados com FS a 0,5 mg/mL apresentaram ótima fluorescência na interface de adesão, enquanto que concentração menor desse corante não produziu sinal suficiente. Padrões morfológicos aparentemente atípicos foram observados na interface de adesão, quando da associação do adesivo SE com o corante FS. A adição de RB e FS nas quatro concentrações indicadas aos adesivos MP e SE não afetou o GC nem o AC em comparação com os grupos de controle correspondentes. Em suma, a RB mostra-se um corante mais versátil que a FS na avaliação morfológica das interfaces dentina-MP e dentina-SE via MCVL. A menor faixa de concentrações de RB nos adesivos MP e SE, na qual é possível produzir imagens satisfatórias das interfaces, situa-se entre 0,10,02 mg/mL. Já o corante FS deve ser adicionado a esses adesivos a pelo menos 0,5 mg/mL para produzir níveis de fluorescência satisfatórios na interface de adesão. A não ocorrência de efeitos deletérios sobre a polimerização e a molhabilidade das resinas estabelece uma margem de segurança para a incorporação desses agentes fluorescentes (em concentração 0,5 mg/mL) nesses sistemas monoméricos. / The addition of fluorescent dyes to dental adhesives makes it possible to investigate the spatial distribution of such resin-based materials in the tooth-restoration interface, using confocal laser scanning microscopy (CLSM). Literature indicates a lack of standardization on the application of fluorescent agents for this purpose. This work presents strategies for adding rhodamine B (RB) and fluorescein sodium salt (FS) to a three-step etch-and-rinse adhesive system, Adper Scotchbond Multi-Purpose (MP), and a two-step self-etching one, Clearfil SE Bond (SE), both regarded as \"gold standard\" in restorative dentistry. The main objectives were (a) to determine the lowest range of RB and FS concentrations required to produce suitable images of the dentin-adhesive interface via CLSM and (b) to investigate potential effects of addition of these dyes on some resin properties. The adhesives were labeled with RB or FS at decreasing concentrations (0.5, 0.1, 0.02 and 0.004 mg/mL) by means of a semi-direct dispersion method. The photophysical/fluorescent behavior of the labeled resins was investigated by photoluminescence spectroscopy and by CLSM. The adhesives were also investigated with regards to the degree of conversion (DC) and contact angle (CA). A two-way ANOVA of adhesive and treatment was conducted on DC and CA separately, followed by Tukeys test. The maximum emission and excitation wavelengths of RB and FS were influenced by the host polymer and the dye concentration in general. The preliminary CLSM of cured adhesive samples, performed with standardized settings, showed that the fluorescent behavior of RB in MP and SE was very similar in the same dye concentration, unlike the behavior of FS, which was lower in the self-etching adhesive for the highest dye concentration. In dentin, the adhesives prepared with RB at the target concentrations of 0.1 and 0.02 mg/mL presented optimal fluorescence; those with 0.004 mg/mL produced poor signal. Adhesives prepared with FS at 0.5 mg/mL presented optimal fluorescence at the bonding interface, whereas lower concentrations of FS did not produce sufficient signal. Atypical morphological features were observed at the bonding interface, when adhesive SE was used with FS. The addition of RB and FS at the four decreasing concentrations to adhesives MP and SE did not affect DC or CA compared to the corresponding controls. In short, RB is more versatile than FS for the morphological characterization of dentin-MP and dentin-SE interfaces via MCVL. The lowest range of RB concentrations in adhesives MP and SE that can produce suitable images of the bonding interface lies between 0.10.02 mg/mL. The dye FS should be added to these adhesives at 0.5 mg/mL at least to produce satisfactory fluorescence levels at the bonding interface. Since negative effects on polymerization and wettability of the resins were not observed, the use of RB and FS (in concentration 0.5 mg/mL) together with MP and SE should be reliable in terms of resin properties.

This study explores the application of two methods of spectroscopy; Near Infrared spectroscopy (NIR) and Fourier transform spectroscopy (FTIR) as alternative approaches for measuring glucocorticoid metabolites in chimpanzee feces. The goals of this study were twofold: The first was to determine if cortisol can be identified within the NIR and/or FTIR spectra of chimpanzee fecal hormone extract in ethanol solution. The second objective was to determine the capability of NIR and FTIR to predict FGM concentrations obtained using standard laboratory methods. Fecal glucocorticoid concentrations measured by Enzyme Immunoassay were used as the reference data of partial least square (PLS) regression of fecal extract NIR spectra and FTIR spectra. Low accuracies (NIR: R2 = 0.152; FTIR: R2 = 0.199) were obtained from regression models using data from both methods. Though this study did not successfully demonstrate the feasibility of using NIR and FTIR to qualify and quantify FGMs, it is likely not a reflection of the capabilities of the technology, but rather of appropriate sample types and preparation methods. / Includes bibliography. / Thesis (M.A.)--Florida Atlantic University, 2016. / FAU Electronic Theses and Dissertations Collection

Thesis (Master of Applied Sciences in Chemistry)--Cape Peninsula University of Technology, 2018. / Biosensors have been on the forefront to provide clinical diagnosis tools for various diseases. Proper selection of biomarkers as well as chemical electrode modification is key in the fabrication of electrochemical biosensors. Hence, electrode modified with nanomaterials devices to improve electroanalytical applications. These nanomaterials were functionalized to improve conductivity, accelerate signal transduction and amplify biorecognition events. Thus, resulting in novel sensing platforms that are highly sensitive and selective towards the target analyte. In this study, gold nanorods (Au NRs) capped with CTAB, zeolitic imidazole framework were synthesised using the seed mediated and hydrothermal method respectively. Composites of gold nanorods with cysteine, ZIF-8 or both were also synthesised. All synthesised materials were characterized using ultraviolet–visible (UV-Vis) spectroscopy, Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), X-Ray diffraction (XRD) and cyclic voltammetry (CV) techniques. The obtained results confirmed the synthesis of the nanomaterials and composites. Identification of the ideal platform for fabrication of a transducer with the best electrochemical response was determined by studying the combinations of the synthesised nanomaterials and composites. The studied parameters were surface coverage, conductivity, rate of electron transfer constant. Cysteine-Au NRs composites platforms, had exceptional properties hence its synthesis optimisation of was undertaken. The effect of CTAB, reaction time, volume and concentration ratio of Au:Cysteine, temperature and pH on the composite properties were assessed. However, this composite’s electrochemical properties decreased when bioconjugated with the antibodies. Hence, the choice of Au NRs CTAB functionalised ZIF-8 (CTABAu/ZIF-8) as the transducer for biosensor applications due to a more favourable biocompatibility. Biosensor fabrication was done by drop coating glassy carbon electrode with the CTABAu/ZIF-8 forming a transducer followed by immobilisation of the antibody (Ab) using a covalent attachment method with glutaraldehyde (GA) as a cross linker. The target analyte, epidermal growth factor (EGF) was interacted with the Ab binding sites via electrostatic forces. All the fabrication steps were optimized for biosensor components, immobilization technique (drop coating and immersion), concentration and incubation time of linker and bioreceptor, as well as the synthesis of the CTABAu-ZIF-8 composites where in situ and ex situ techniques were compared together with the effect of the concentration ratio of Au: ZIF-8. There was also an analysis of optimum pH. Optimum conditions were found to be immersion in 3 % GA and 2 μg/ml Ab, with incubation times of 8, 10 and 5 minutes for GA, Ab and EGF respectively at a pH of 6. The following electroanalytical techniques: cyclic voltammetry (CV), differential pulse voltammetry (DPV) and square wave voltammetry (SWV) were utilised for EGF detection. The DPV showed better reproducibility, higher currents and better resolution hence; it was the method of choice. The technique’s optimisation involved assessments of the effect of step potential, starting potential and pulse amplitude. The optimum response for pulse amplitude, step potential and starting potential were 60 mV, 20 mV and 0.5 V respectively. The biosensor analytical parameters were linear towards EGF in the concentration range from 2 to 100 nM with a detection limit of 0.58 nM. Reproducibility and repeatability tests were acceptable, and the biosensor had a stability over 80 % within 15 days. There was no interference observed in the presence of glucose and creatine. The EGF biosensor was successfully applied in urine and saliva analysis, obtaining 67.5 and 3.12 nM respectively. This biosensor’s positive outcome strongly suggests its potential as a diagnosis tool for early detection of kidney disease as it was able to detect EGF concentration within physiological levels of EGF in normal kidney function.

A tireoide é uma das principais glândulas do nosso sistema endócrino e responsável pela produção dos hormônios Triodotirosina (T3) e Tetraiodotironina (T4) que são responsáveis pelo controle metabólico basal. A tireoide pode ser acometida por neoplasias benignas e malignas que pode levar a uma produção anormal de hormônios e levando a sérios problemas de saúde. O diagnóstico de algumas destas neoplasias por citologia ainda não possuem altas taxas de sensibilidade e especificidade para os casos com padrão folicular, portanto o desenvolvimento de novos métodos que se baseiam na analise de características bioquímicas dos tecidos tireoidianos se faz necessário. O presente trabalho caracterizou tecidos tireoidianos normais e com bócio nodular por meio da espectroscopia no infravermelho. As imagens espectrais foram adquiridas com alta resolução e aliadas a um grande processamento das mesmas utilizando métodos de estatística multivariada se obteve diferenças bioquímicas dos tecidos. Nos resultados obtidos no presente trabalho, a maior diferença observada está na região da amida I que está relacionada a estrutura secundária da tiroglobulina devido ao processo de incorporação de iodo. Outras observações que foram feitas são diferenças em glicosilaçao do tipo N. A partir dos resultados obtidos nesta tese conclui-se que a técnica de imageamento por microespectroscopia no infravermelho é capaz de observar diferenças bioquímicas importantes entre tecidos tireoidianos sadios e com bócio, apresentando grande potencial para o desenvolvimento no futuro de novas metodologias, baseadas em espectroscopia vibracional. / The thyroid is one of the major glands of endocrine system; it is responsible for the production triodothyronine(T3) and thyroxine (T4) hormones, which are responsible for metabolic basal control. The thyroid can be affected by benign and malignant lesions, that leads to increased hormone levels (hyperthyroidism) or reduce their production (hypothyroidism). This abnormal hormones production by the thyroid gland can lead to serious health problems. The present cytological diagnosis for these tumors do not present high level of sensitivity and specificity, thus the development of new methods based on the analysis of biochemical characteristics of the thyroid tissue is required. This study used the infrared spectroscopy to characterize normal thyroid tissue or tissue with nodular goiter. The spectral images were obtained with high resolution, and multivariate statistics analysis were used to observe biochemical tissues differences. The most remarkable difference observed was at the amide I region, which is related to the secondary structure of thyroglobulin, due to the process of incorporation of iodine. It was also observed differences in the N glycosylation. From the results of this study, it was possible conclude that the technique of infrared microspectroscopy and imaging is able to observe important biochemical differences between healthy thyroid tissue and goiter.

A tireoide é uma das principais glândulas do nosso sistema endócrino e responsável pela produção dos hormônios Triodotirosina (T3) e Tetraiodotironina (T4) que são responsáveis pelo controle metabólico basal. A tireoide pode ser acometida por neoplasias benignas e malignas que pode levar a uma produção anormal de hormônios e levando a sérios problemas de saúde. O diagnóstico de algumas destas neoplasias por citologia ainda não possuem altas taxas de sensibilidade e especificidade para os casos com padrão folicular, portanto o desenvolvimento de novos métodos que se baseiam na analise de características bioquímicas dos tecidos tireoidianos se faz necessário. O presente trabalho caracterizou tecidos tireoidianos normais e com bócio nodular por meio da espectroscopia no infravermelho. As imagens espectrais foram adquiridas com alta resolução e aliadas a um grande processamento das mesmas utilizando métodos de estatística multivariada se obteve diferenças bioquímicas dos tecidos. Nos resultados obtidos no presente trabalho, a maior diferença observada está na região da amida I que está relacionada a estrutura secundária da tiroglobulina devido ao processo de incorporação de iodo. Outras observações que foram feitas são diferenças em glicosilaçao do tipo N. A partir dos resultados obtidos nesta tese conclui-se que a técnica de imageamento por microespectroscopia no infravermelho é capaz de observar diferenças bioquímicas importantes entre tecidos tireoidianos sadios e com bócio, apresentando grande potencial para o desenvolvimento no futuro de novas metodologias, baseadas em espectroscopia vibracional. / The thyroid is one of the major glands of endocrine system; it is responsible for the production triodothyronine(T3) and thyroxine (T4) hormones, which are responsible for metabolic basal control. The thyroid can be affected by benign and malignant lesions, that leads to increased hormone levels (hyperthyroidism) or reduce their production (hypothyroidism). This abnormal hormones production by the thyroid gland can lead to serious health problems. The present cytological diagnosis for these tumors do not present high level of sensitivity and specificity, thus the development of new methods based on the analysis of biochemical characteristics of the thyroid tissue is required. This study used the infrared spectroscopy to characterize normal thyroid tissue or tissue with nodular goiter. The spectral images were obtained with high resolution, and multivariate statistics analysis were used to observe biochemical tissues differences. The most remarkable difference observed was at the amide I region, which is related to the secondary structure of thyroglobulin, due to the process of incorporation of iodine. It was also observed differences in the N glycosylation. From the results of this study, it was possible conclude that the technique of infrared microspectroscopy and imaging is able to observe important biochemical differences between healthy thyroid tissue and goiter.