Exploring Forensic Nursing Ethics and Practice: Roles, Loyalties and Photodocumentation Practices

Thesis (Ph.D.)--University of Washington, 2016-08

Evaluation of Select Food Additive Exposures in Children with Crohn's Disease

Thesis (Master's)--University of Washington, 2016-09

Avaliação da viabilidade da aplicação de técnicas biomoleculares na autenticação de produtos alimentares

A carne continua a ser a fonte proteica mais comum no quotidiano das pessoas. Além disso, os produtos cárneos processados apresentam-se como uma mais-valia nas suas vidas agitadas. Este tipo de produto torna difícil a diferenciação das carnes utilizadas na sua confecção, sendo por isso propícios a adulteração. A Reacção em Cadeia da Polimerase (PCR) tem ganho cada vez mais importância nos laboratórios de biologia molecular, revelando-se uma técnica de análise rápida, sensível e altamente específica na identificação de espécies em produtos alimentares. No entanto, vários factores podem interferir com o processo de amplificação, pelo que alguns cuidados devem ser implementados desde a aquisição da amostra a analisar, ao seu acondicionamento e posterior extração de ADN. Existem inúmeros protocolos de extração de ADN, devendo para cada estudo avaliar-se e optar-se pelo mais adequado, considerando a finalidade estabelecida para a amostra extraída. O trabalho laboratorial apresentado nesta dissertação baseou-se em três etapas principais. Inicialmente, avaliaram-se diferentes protocolos de extração de ADN, utilizando-se amostras de carne adquiridas num talho. Entre os protocolos testados, o método de Brometo de Cetil-Trimetil-Amónio (CTAB) modificado foi o que permitiu obter amostras de ADN com maior concentração e elevado nível de pureza. Posteriormente, foram testados e optimizados diferentes protocolos de amplificação, por PCR em tempo real, para a detecção das espécies Bos taurus (vaca), Sus scrofa (porco), Equus caballus (cavalo) e Ovis aries (ovelha). Foram empregues primers específicos de espécie para a detecção de genes mitocondriais e genómicos, consoante cada protocolo. Para o caso concreto do porco, foi efectuada a avaliação de dois protocolos, singleplex com EvaGreen® e tetraplex com AllHorse, para possível aplicação dos mesmos na sua quantificação. Os resultados demonstraram elevada especificidade e sensibilidade das reacções para esta espécie, permitindo a sua detecção até um limite de 0,001 ng e 0,1%, respectivamente. Somente a primeira metodologia se mostrou adequada para quantificação. Por último, as metodologias sugeridas foram aplicadas com sucesso na análise de 4 amostras comerciais de hambúrgueres, tendo-se verificado a consistência da rotulagem em todos os casos, no que concerne a composição em termos de espécies animais. O interesse de trabalhos neste âmbito recai na importância da autenticidade dos rótulos de produtos alimentares, principalmente nos produtos cárneos, para segurança dos consumidores e salvaguarda dos produtores.

Strävan efter värdighet : En studie av servicearbete

The focus of this study is dignity in low status service work. Using labels such as bad jobs, McJobs and dirty work, these jobs have been described as low-skilled, low-paid, monotonous and physically demanding with lack of voice and no job security. Research on dignity at work is especially relevant in a time when different ambitions for more dignified work, initiated by political parties as well as unions, tend to be forgotten or down-prioritized. This study investigates what conditions are preventing dignity among low status service workers and how they create and maintain dignity for themselves. What briefly has been found is that dignity can be prevented by unreasonable demands, constant control, exposed work and mismanagement. Moreover, customerprerogative can prevent dignity when employees are being mistreated by disrespectful customers. Dignity is also hindered by frightening customers, especially in the case of sexual harassment, threats and violence. In this study theories about working conditions and professional status are brought together to explain experiences of dignity at work. Service workers do not only have managers to deal with, but also customers whose treatment is reflected by the status of the service occupation. Besides, working conditions and professional status are two mechanisms acting together when it comes to experiences of dignity at work and may thus result in double tensions in daily work.   Acts for dignity, meaning different ways in which the service workers create and maintain dignity for themselves, are reactions to the obstacles to dignity at work. Three different categories of acts for dignity can be found. The identity-bolstering acts help the workers maintain their professional identity or self-image when it is threatened by different obstacles to dignity. The justifying acts mean that the workers legitimize different obstacles to dignity. Finally, the compensating acts help the workers to even out different obstacles to dignity.

La réception du hip-hop chez des rappeurs afro-québécois dans la ville de Québec : appropriation intersectionnelle de problématiques multidimensionnelles

Cette thèse analyse les rappeurs afro-québécois des communautés de Limoilou (Limoilou Starz) et leurs amis de Montcalm comme des acteurs qui s’approprient des éléments identitaires, de lutte sociale et de survie économique, issus des problématiques globales et américaines du hip-hop, au service de leurs défis particuliers. Les divers mérites des luttes menées au moyen de leur art sont acquis par des formes spécifiques de capital. Ainsi les moyens utilisés par les rappeurs sont principalement les paroles de chansons, les prises de parole publique, dans les médias et sur scène, les campagnes d’affichage, l’utilisation des réseaux numériques, l’entreprise économique autonome (photographie, vidéographie, gestion des artistes, vente de vêtements). Ces moyens spécifiques se rattachent à d’autres principes et actions non explicités et sociohistoriquement ancrés. À partir d’une enquête ethnographique menée auprès de 31 participants dans la ville de Québec, j’utilise le concept de « réception différenciée » (Hall, 1980; Morley, 1980) pour décrire le processus de résistance des différents pratiquants et entrepreneurs de la musique aux dominations provenant de groupes divers. Trois principaux groupes de domination sont examinés : les agents d’institutions étatiques (comme les policiers du Service de Police de la Ville de Québec), les agents d’entreprises privées (comme les patrons de grandes boites de nuit et les propriétaires de labels musicaux indépendants) et les groupes et individus du milieu hip-hop, à travers leurs stratégies d’intimidation. La théorie « émergente » ou emergent-fit (Guillemette, 2006; Guillemette et Luckerhoff, 2009) permet d’entrevoir la musique hip-hop en amont comme une structure multidimensionnelle (sociale, identitaire, politique et économique) et intersectionnelle (intersection de plusieurs catégories interreliées, relatives au lieu de résidence, à la race et aux capacités économiques), et en aval comme un champ musical (Bourdieu, 1976 et 1989; Rimmer, 2010) renégocié. Cette structure a pris forme et s’est transformée grâce aux dispositions mentales et physiques (habitus) des acteurs étudiés. Les résultats de cette recherche montrent que certains rappeurs et leurs autres collègues artistes hip-hop— ainsi que quelques entrepreneurs— résistent à plusieurs sortes de domination. D’autres encore acceptent ces dominations sous forme d’idéologies, même en le reconnaissant explicitement. Par contre, une infime partie des acteurs étudiés les rejettent complètement. Ainsi, l’appropriation multidimensionnelle et intersectionnelle des sens dominants à travers le hip-hop mène à plusieurs formes de lecture de la domination et de la résistance.

Security and Trust in Distributed Computation

We propose three research problems to explore the relations between trust and security in the setting of distributed computation. In the first problem, we study trust-based adversary detection in distributed consensus computation. The adversaries we consider behave arbitrarily disobeying the consensus protocol. We propose a trust-based consensus algorithm with local and global trust evaluations. The algorithm can be abstracted using a two-layer structure with the top layer running a trust-based consensus algorithm and the bottom layer as a subroutine executing a global trust update scheme. We utilize a set of pre-trusted nodes, headers, to propagate local trust opinions throughout the network. This two-layer framework is flexible in that it can be easily extensible to contain more complicated decision rules, and global trust schemes. The first problem assumes that normal nodes are homogeneous, i.e. it is guaranteed that a normal node always behaves as it is programmed. In the second and third problems however, we assume that nodes are heterogeneous, i.e, given a task, the probability that a node generates a correct answer varies from node to node. The adversaries considered in these two problems are workers from the open crowd who are either investing little efforts in the tasks assigned to them or intentionally give wrong answers to questions. In the second part of the thesis, we consider a typical crowdsourcing task that aggregates input from multiple workers as a problem in information fusion. To cope with the issue of noisy and sometimes malicious input from workers, trust is used to model workers' expertise. In a multi-domain knowledge learning task, however, using scalar-valued trust to model a worker's performance is not sufficient to reflect the worker's trustworthiness in each of the domains. To address this issue, we propose a probabilistic model to jointly infer multi-dimensional trust of workers, multi-domain properties of questions, and true labels of questions. Our model is very flexible and extensible to incorporate metadata associated with questions. To show that, we further propose two extended models, one of which handles input tasks with real-valued features and the other handles tasks with text features by incorporating topic models. Our models can effectively recover trust vectors of workers, which can be very useful in task assignment adaptive to workers' trust in the future. These results can be applied for fusion of information from multiple data sources like sensors, human input, machine learning results, or a hybrid of them. In the second subproblem, we address crowdsourcing with adversaries under logical constraints. We observe that questions are often not independent in real life applications. Instead, there are logical relations between them. Similarly, workers that provide answers are not independent of each other either. Answers given by workers with similar attributes tend to be correlated. Therefore, we propose a novel unified graphical model consisting of two layers. The top layer encodes domain knowledge which allows users to express logical relations using first-order logic rules and the bottom layer encodes a traditional crowdsourcing graphical model. Our model can be seen as a generalized probabilistic soft logic framework that encodes both logical relations and probabilistic dependencies. To solve the collective inference problem efficiently, we have devised a scalable joint inference algorithm based on the alternating direction method of multipliers. The third part of the thesis considers the problem of optimal assignment under budget constraints when workers are unreliable and sometimes malicious. In a real crowdsourcing market, each answer obtained from a worker incurs cost. The cost is associated with both the level of trustworthiness of workers and the difficulty of tasks. Typically, access to expert-level (more trustworthy) workers is more expensive than to average crowd and completion of a challenging task is more costly than a click-away question. In this problem, we address the problem of optimal assignment of heterogeneous tasks to workers of varying trust levels with budget constraints. Specifically, we design a trust-aware task allocation algorithm that takes as inputs the estimated trust of workers and pre-set budget, and outputs the optimal assignment of tasks to workers. We derive the bound of total error probability that relates to budget, trustworthiness of crowds, and costs of obtaining labels from crowds naturally. Higher budget, more trustworthy crowds, and less costly jobs result in a lower theoretical bound. Our allocation scheme does not depend on the specific design of the trust evaluation component. Therefore, it can be combined with generic trust evaluation algorithms.

Assessment of the quality of sample labelling for clinical research

Objective: To assess the quality of the labels for clinical trial samples through current regulations, and to analyze its potential correlation with the specific characteristics of each sample. Method: A transversal multicenter study where the clinical trial samples from two third level hospitals were analyzed. The eleven items from Directive 2003/94/EC, as well as the name of the clinical trial and the dose on the label cover, were considered variables for labelling quality. The influence of the characteristics of each sample on labelling quality was also analyzed. Outcome: The study included 503 samples from 220 clinical trials. The mean quality of labelling, understood as the proportion of items from Appendix 13, was of 91.9%. Out of these, 6.6% did not include the name of the sample in the outer face of the label, while in 9.7% the dose was missing. The samples with clinical trial-type samples presented a higher quality (p < 0.049), blinding reduced their quality (p = 0.017), and identification by kit number or by patient increased it (p < 0.01). The promoter was the variable which introduced the highest variability into the analysis. Conclusions: The mean quality of labelling is adequate in the majority of clinical trial samples. The lack of essential information in some samples, such as the clinical trial code and the period of validity, is alarming and might be the potential source for dispensing or administration errors.

Contextual text mining

With the dramatic growth of text information, there is an increasing need for powerful text mining systems that can automatically discover useful knowledge from text. Text is generally associated with all kinds of contextual information. Those contexts can be explicit, such as the time and the location where a blog article is written, and the author(s) of a biomedical publication, or implicit, such as the positive or negative sentiment that an author had when she wrote a product review; there may also be complex context such as the social network of the authors. Many applications require analysis of topic patterns over different contexts. For instance, analysis of search logs in the context of the user can reveal how we can improve the quality of a search engine by optimizing the search results according to particular users; analysis of customer reviews in the context of positive and negative sentiments can help the user summarize public opinions about a product; analysis of blogs or scientific publications in the context of a social network can facilitate discovery of more meaningful topical communities. Since context information significantly affects the choices of topics and language made by authors, in general, it is very important to incorporate it into analyzing and mining text data. In general, modeling the context in text, discovering contextual patterns of language units and topics from text, a general task which we refer to as Contextual Text Mining, has widespread applications in text mining. In this thesis, we provide a novel and systematic study of contextual text mining, which is a new paradigm of text mining treating context information as the ``first-class citizen.'' We formally define the problem of contextual text mining and its basic tasks, and propose a general framework for contextual text mining based on generative modeling of text. This conceptual framework provides general guidance on text mining problems with context information and can be instantiated into many real tasks, including the general problem of contextual topic analysis. We formally present a functional framework for contextual topic analysis, with a general contextual topic model and its various versions, which can effectively solve the text mining problems in a lot of real world applications. We further introduce general components of contextual topic analysis, by adding priors to contextual topic models to incorporate prior knowledge, regularizing contextual topic models with dependency structure of context, and postprocessing contextual patterns to extract refined patterns. The refinements on the general contextual topic model naturally lead to a variety of probabilistic models which incorporate different types of context and various assumptions and constraints. These special versions of the contextual topic model are proved effective in a variety of real applications involving topics and explicit contexts, implicit contexts, and complex contexts. We then introduce a postprocessing procedure for contextual patterns, by generating meaningful labels for multinomial context models. This method provides a general way to interpret text mining results for real users. By applying contextual text mining in the ``context'' of other text information management tasks, including ad hoc text retrieval and web search, we further prove the effectiveness of contextual text mining techniques in a quantitative way with large scale datasets. The framework of contextual text mining not only unifies many explorations of text analysis with context information, but also opens up many new possibilities for future research directions in text mining.

Between Bacchus and Dionysus: European Union policies on rural development: the cases of Nemea and Basto wine regions

A actividade vitivinícola possui um conjunto diverso de características presentes no solo, território e comunidade que fazem parte do património cultural de uma determinada região. Quando a tradição se traduz num conceito como terroir que é formado por características territoriais, sociais e culturais de uma região rural, o vinho apresenta uma “assinatura” que se escreve “naturalmente” no paladar regionalmente identificado. Os vinhos da Região de Nemea, na Grécia e de Basto (Região dos Vinhos Verdes) em Portugal, estão ambos sob a proteção dos regulamentos das Denominações de Origem. No entanto, apesar de ambos serem regulados por sistemas institucionais de certificação e controlo de qualidade, afigura-se a necessidade de questionar se o património cultural e a identidade territorial específica, “impressa” em ambos os terroirs, pode ser protegida num sentido mais abrangente do que apenas origem e qualidade. Em Nemea, a discussão entre os produtores diz respeito ao estabelecimento de sub-zonas, isto é incluir na regulação PDO uma diferente categorização territorial com base no terroir. Ou seja, para além de estar presente no rótulo a designação PDO, as garrafas incluirão ainda informação certificada sobre a área específica (dentro do mesmo terroir) onde o vinho foi produzido. A acontecer resultaria em diferentes status de qualidade de acordo com as diferentes aldeias de Nemea onde as vinhas estão localizadas. O que teria possíveis impactos no valor das propriedades e no uso dos solos. Para além disso, a não participação da Cooperativa de Nemea na SON (a associação local de produtores de vinho) e como tal na discussão principal sobre as mudanças e os desafios sobre o terroir de Nemea constitui um problema no sector vitivinícola de Nemea. Em primeiro lugar estabelece uma relação de não-comunicação entre os dois mais importantes agentes desse sector – as companhias vinícolas e a Cooperativa. Em segundo lugar porque constituiu uma possibilidade real, não só para os viticultores ficarem arredados dessa discussão, como também (porque não representados pela cooperativa) ficar impossibilitado um consenso sobre as mudanças discutidas. Isto poderá criar um ‘clima’ de desconfiança levando a discussão para ‘arenas’ deslocalizadas e como tal para decisões ‘desterritorializadas’ Em Basto, há vários produtores que começaram a vender a sua produção para distribuidoras localizadas externamente à sub-região de Basto, mas dentro da Região dos Vinhos Verdes, uma vez que essas companhias tem um melhor estatuto nacional e internacional e uma melhor rede de exportações. Isto está ainda relacionado com uma competição por uma melhor rede de contactos e status mais forte, tornando as discussões sobre estratégias comuns para o desenvolvimento rural e regional de Basto mais difícil de acontecer (sobre isto a palavra impossível foi constantemente usada durante as entrevistas com os produtores de vinho). A relação predominante entre produtores é caracterizada por relações individualistas. Contudo foi observado que essas posições são ainda caracterizadas por uma desconfiança no interior da rede interprofissional local: conflitos para conseguir os mesmos potenciais clientes; comprar uvas a viticultores com melhor rácio qualidade/preço; estratégias individuais para conseguir um melhor status político na relação com a Comissão dos Vinhos Verdes. Para além disso a inexistência de uma activa intermediação institucional (autoridades municipais e a Comissão de Vinho Verde), a inexistência entre os produtores de Basto de uma associação ou mesmo a inexistência de uma cooperativa local tem levado a região de Basto a uma posição de subpromoção nas estratégias de promoção do Vinho Verde em comparação com outras sub-regiões. É também evidente pelos resultados que as mudanças no sector vitivinícolas na região de Basto têm sido estimuladas de fora da região (em resposta também às necessidades dos mercados internacionais) e raramente de dentro – mais uma vez, ‘arenas’ não localizadas e como tal decisões desterritorializadas. Nesse sentido, toda essa discussão e planeamento estratégico, terão um papel vital na preservação da identidade localizada do terroir perante os riscos de descaracterização e desterritorialização. Em suma, para ambos os casos, um dos maiores desafios parece ser como preservar o terroir vitivinícola e como tal o seu carácter e identidade local, quando a rede interprofissional em ambas as regiões se caracteriza, tanto por relações não-consensuais em Nemea como pelo modus operandi de isolamento sem comunicação em Basto. Como tal há uma necessidade de envolvimento entre os diversos agentes e as autoridades locais no sentido de uma rede localizada de governança. Assim sendo, em ambas as regiões, a existência dessa rede é essencial para prevenir os efeitos negativos na identidade do produto e na sua produção. Uma estratégia de planeamento integrado para o sector será vital para preservar essa identidade, prevenindo a sua desterritorialização através de uma restruturação do conhecimento tradicional em simultâneo com a democratização do acesso ao conhecimento das técnicas modernas de produção vitivinícola.

Inclusion of salt form on prescription medication labeling as a source of patient confusion: a pilot study

Background: It has been estimated that 10,000 patient injuries occur in the US annually due to confusion involving drug names. An unexplored source of patient misunderstandings may be medication salt forms. Objective: The objective of this study was to assess patient knowledge and comprehension regarding the salt forms of medications as a potential source of medication errors. Methods: A 12 item questionnaire which assessed patient knowledge of medication names on prescription labels was administered to a convenience sample of patients presenting to a family practice clinic. Descriptive statistics were calculated and multivariate analyses were performed. Results: There were 308 responses. Overall, 41% of patients agreed they find their medication names confusing. Participants correctly answered to salt form questions between 12.1% and 56.9% of the time. Taking more prescription medications and higher education level were positively associated with providing more correct answers to 3 medication salt form knowledge questions, while age was negatively associated. Conclusions: Patient misconceptions about medication salt forms are common. These findings support recommendations to standardize the inclusion or exclusion of salt forms. Increasing patient education is another possible approach to reducing confusion.

RICH AND EFFICIENT VISUAL DATA REPRESENTATION

Increasing the size of training data in many computer vision tasks has shown to be very effective. Using large scale image datasets (e.g. ImageNet) with simple learning techniques (e.g. linear classifiers) one can achieve state-of-the-art performance in object recognition compared to sophisticated learning techniques on smaller image sets. Semantic search on visual data has become very popular. There are billions of images on the internet and the number is increasing every day. Dealing with large scale image sets is intense per se. They take a significant amount of memory that makes it impossible to process the images with complex algorithms on single CPU machines. Finding an efficient image representation can be a key to attack this problem. A representation being efficient is not enough for image understanding. It should be comprehensive and rich in carrying semantic information. In this proposal we develop an approach to computing binary codes that provide a rich and efficient image representation. We demonstrate several tasks in which binary features can be very effective. We show how binary features can speed up large scale image classification. We present learning techniques to learn the binary features from supervised image set (With different types of semantic supervision; class labels, textual descriptions). We propose several problems that are very important in finding and using efficient image representation.

A rewarding experience? Exploring how crowdfunding is affecting music industry business models.

This paper provides an exploratory study of how rewards-based crowdfunding affects business model development for music industry artists, labels and live sector companies. The empirical methodology incorporated a qualitative, semi-structured, three-stage interview design with fifty seven senior executives from industry crowdfunding platforms and three stakeholder groups. The results and analysis cover new research ground and provide conceptual models to develop theoretical foundations for further research in this field. The findings indicate that the financial model benefits of crowdfunding for independent artists are dependent on fan base demographic variables relating to age group and genre due to sustained apprehension from younger audiences. Furthermore, major labels are now considering a more user-centric financial model as an innovation strategy, and the impact of crowdfunding on their marketing model may already be initiating its development in terms of creativity, strength and artist relations.

Deep convolutional neural networks for semantic video object segmentation

In this thesis, we propose to infer pixel-level labelling in video by utilising only object category information, exploiting the intrinsic structure of video data. Our motivation is the observation that image-level labels are much more easily to be acquired than pixel-level labels, and it is natural to find a link between the image level recognition and pixel level classification in video data, which would transfer learned recognition models from one domain to the other one. To this end, this thesis proposes two domain adaptation approaches to adapt the deep convolutional neural network (CNN) image recognition model trained from labelled image data to the target domain exploiting both semantic evidence learned from CNN, and the intrinsic structures of unlabelled video data. Our proposed approaches explicitly model and compensate for the domain adaptation from the source domain to the target domain which in turn underpins a robust semantic object segmentation method for natural videos. We demonstrate the superior performance of our methods by presenting extensive evaluations on challenging datasets comparing with the state-of-the-art methods.

Chemical characterization and in vitro cyto- and genotoxicity of ‘legal high’ products containing Kratom (Mitragyna speciosa)

Kratom is a popular ‘legal high’ mainly constituted by alkaloids extracted from the Mitragyna speciosa plant with mitragynine (MG) as the dominant active substance. The increasing use of Kratom for recreational purposes has alerted risk assessment bodies of the lack of information on the real composition and its potential health risks. The present study aimed to determine and compare the MG composition of 13 commercial products of Kratom sold online and in “smartshops”, by gas chromatography–mass spectrometry. For the first time, the cytotoxicity induced by pure MG and Kratom, extracts was evaluated in in vitro models of human intestinal (Caco-2) and neuronal (SH-SY5Y) cells after 6 and 24 h. Genotoxicity was also evaluated in intestinal Caco-2 cells following 24 h of exposure to subtoxic concentrations using the comet assay. The obtained results revealed an inconsistency between the information (‘power’) provided in labels and the MG content. Cytotoxicity tests revealed a concentration-dependent decrease in cell viability in both cellular models, with the SH-SY5Y cells being more sensitive to the Kratom extracts. The resin and the ‘powered extracts’ were the most cytotoxic samples, with IC50 values significantly lower than the leaf extracts and pure MG (P < 0.0001 vs. leaf extracts and MG). In addition, significant DNA damage was observed in Caco-2 cells exposed to these extracts but not to pure MG, which suggests that other substances present in the extracts or interactions involving Kratom components might be responsible for the observed effects.

Análise elementar de águas e sumos

A concentração de seis elementos: Cd, Cr, Pb, Ni, Zn e Fe foi medida em sessenta e sete amostras de sumos de fruta 100 %, duas amostras de refrigerantes, dez amostras de concentrados de sumos e sete amostras de águas de diluição utilizadas no processamento dos sumos. As amostras de sumos representam numa prespectiva bastante abrangente o mercado Português de sumos de fruta 100 %. Os refrigerantes concentrados e águas de diluição foram cedidos por duas empresas fabricantes de sumos Portuguesas. As concentrações elementares foram medidas pelas técnicas de FAAS e GFAAS e foi medido também o grau Brix dos sumos. Os factores: fruta, percentagem de fruta, origem, agricultura, tratamento, embalagem, conservação e processo foram obtidos por informação do fabricante nos rótulos dos produtos e por contacto directo. Caracterizou-se o mercado em termos da concentração desses elementos e caracterizou-se a sua diluição comparando-a com valores de referência do mercado Europeu. Mediu-se o grau de associação entre os diversos parâmetros e a concentração final elementar dos sumos e utilizou-se a análise de agrupamentos, a análise de correspondência múltipla e a análise factorial para reestruturar a matriz de dados. Dos resultados obtidos, os sumos de fruta apresentam a seguinte ordem de grandeza nas suas concentrações elementares: Cd