Automatic musical instrument recognition for multimedia indexing

Trabalho apresentado no âmbito do Mestrado em Engenharia Informática, como requisito parcial para obtenção do grau de Mestre em Engenharia Informática

Propriedades Psicométricas da Versão Portuguesa do Kansas City Cardiomyopathy Questionnaire na Miocardiopatia Dilatada com Insuficiência Cardíaca Congestiva

Several studies have shown that patients with congestive heart failure (CHF) have a compromised health-related quality of life (HRQL), and this, in recent years, has become a primary endpoint when considering the impact of treatment of chronic conditions such as CHF. OBJECTIVES: To evaluate the psychometric properties of the Portuguese version of a new specific instrument to measure HRQL in patients hospitalized for CHF: the Kansas City Cardiomyopathy Questionnaire (KCCQ). METHODS: The KCCQ was applied to a sample of 193 consecutive patients hospitalized for CHF. Of these, 105 repeated the assessment 3 months after admission, with no events during this period. Mean age was 64.4 +/- 12.4 years (21-88), and 72.5% were 72.5% male. CHF was of ischemic etiology in 4% of cases. RESULTS: This version of the KCCQ was subjected to statistical validation, with assessment of reliability and validity, similar to the American version. Reliability was assessed by the internal consistency of the domains and summary scores, which showed similar values of Cronbach alpha (0.50-0.94). Validity was assessed by convergence, sensitivity to differences between groups and sensitivity to changes in clinical condition. We evaluated the convergent validity of all domains related to functionality, through the relationship between them and a measure of functionality, the New York Heart Association (NYHA) classification. Significant correlations were found (p < 0.01) for this measure of functionality i patients with CHF. Analysis of variance between the physical limitation domain, the summary scores and NYHA class was performed and statistically significant differences were found (F = 23.4; F = 36.4; F = 37.4, p = 0.0001) in the ability to discriminate severity of clinical condition. A second evaluation was performed on 105 patients at the 3-month follow-up outpatient appointment, and significant changes were observed in the mean scores of the domains assessed between hospital admission and the clinic appointment (differences from 14.9 to 30.6 on a scale of 0-100), indicating that the domains assessed are sensitive to changes in clinical condition. The correlation between dimensions of quality of life in the KCCQ is moderate, suggesting that the dimensions are independent, supporting the multifactorial nature of HRQL and the suitability of this measure for its evaluation. CONCLUSION: The KCCQ is a valid instrument, sensitive to change and a specific measure of HRQL in a population with dilated cardiomyopathy and CHF.

Tradução com adaptação cultural para portugês e determinação da reprodutibilidade do questionnaire on respiratory symptoms in preschool children

RESUMO: Introdução: A asma brônquica é uma entidade frequente em idade pediátrica, apresentando uma grande heterogeneidade clínica e significativa morbilidade quando não controlada. A identificação de crianças sintomáticas pode atrasar ou até mesmo diminuir a ocorrência de algumas alterações estruturais. Reconhece-se a necessidade de questionários sobre sintomas respiratórios em língua portuguesa, devidamente validados, que tenham como população-alvo os grupos etários inferiores a 3 anos. Deste modo, será possível não só um conhecimento mais rigoroso da asma e da sibilância infantil mas também a uniformização de metodologias para o desenvolvimento de estratégias a nível nacional. Objetivos: Tradução com adaptação cultural para português e determinação da reprodutibilidade do Questionnaire on respiratory symptoms in preschool children de Strippoli e colaboradores. Material e métodos: A escolha do questionário obedeceu a vários critérios, entre os quais o grupo etário, o tipo e número de perguntas. O Questionnaire on respiratory symptoms in preschool children de Strippoli e colaboradores é um questionário de autopreenchimento, dirigido a crianças entre os 12 e os 24 meses de idade e destinado a estudos epidemiológicos ao nível da comunidade. Aborda aspetos referentes a sintomas respiratórios (sibilância, tosse crónica, sintomas das vias aéreas superiores), cuidados médicos, terapêutica, características ambientais, história familiar e situação social. Procedemos à sua tradução, com especial atenção para a adaptação do ponto de vista cultural e linguístico, utilizando o método da tradução / retroversão, amplamente utilizado e descrito na literatura internacional. Seguidamente determinámos a reprodutibilidade da versão final em língua portuguesa – Questionário de sintomas respiratórios em idade pré-escolar – utilizando o teste-reteste. Para tal, incluíram-se crianças entre os 12 e os 36 meses de idade recrutadas num Centro de Saúde e em creches de Lisboa. A distribuição dos questionários decorreu em duas fases: na primeira fase foram entregues pessoalmente nos locais de recrutamento e na segunda fase foram enviados por correio para os domicílios das crianças, respeitando-se um intervalo mínimo de 2 semanas entre ambos. Resultados: Na primeira fase foram distribuídos 180 questionários, com uma taxa de reposta de 41% (n=74). Na segunda fase enviaram-se para os respetivos domicílios 70 questionários,obtendo-se uma taxa de resposta de 66% (n=46). Para a análise de reprodutibilidade foram incluídos apenas os questionários preenchidos em ambos os momentos pelo mesmo indivíduo (mãe, pai ou representante legal) (n=41). A idade média das crianças foi, na primeira fase, de 22,5 meses e, na segunda fase, de 23,7 meses, com um predomínio do sexo feminino (F:M =1:0,6). A mediana do tempo decorrido entre os dois momentos de preenchimento dos questionários foi de 26 dias. Obtivemos valores de concordância globalmente bons a muito bons, à semelhança do sucedido no trabalho original. Conclusões: Procedemos à tradução e avaliação da reprodutibilidade do Questionnaire on respiratory symptoms in preschool children. Pretende-se que venha a ser uma ferramenta útil para estudos epidemiológicos e programas de rastreio na comunidade, contribuindo deste modo para uma otimização da abordagem da asma / sibilância infantil a nível nacional. -------------ABSTRACT: Background: Asthma is a very common feature in childhood, with important clinical heterogeneity and morbidity if not properly controlled. Identifying symptomatic children may delay or even reduce several structural changes. The development of questionnaires on respiratory symptoms in Portuguese for children under 3 years old will allow not only a more accurate knowledge of infantile asthma and recurrent wheezing but also the standardization of methodologies to develop nationwide strategies. Objectives: The aim of this study was to translate and adapt to the Portuguese culture and to determine the repeatability of the Questionnaire on respiratory symptoms in preschool children by Strippoli et al. Material and methods: The choice of the questionnaire took in consideration several criteria, among which the target age, the type and the number of questions. The Questionnaire on respiratory symptoms in preschool children by Strippoli et al is a parent-completed questionnaire for assessment of respiratory symptoms in 1 to 2-year-old children, developed for cross-sectional and longitudinal studies. It contains sections on respiratory symptoms (wheezing, chronic cough and upper airways symptoms), healthcare utilization, treatment, environmental exposure, family history and social situation. For the process of translation we used the method of translation and back-translation, with particular concern to cultural and linguistic adaptation. To assess the repeatability of the final Portuguese version - Questionário de sintomas respiratórios em idade pré-escolar - we used the test–retest analyses. The questionnaires were distributed to parents of children between 12 and 36 months old attending nurseries and a Primary Care Center of Lisbon. The distribution took place in two phases: the first questionnaires were delivered in person (phase one) and an identical questionnaire was posted to the families that participated in the first phase, 2 weeks after the first one was returned (phase two). Results: The response rates were 41% (180/74) in the first phase and 66% (70/46) in the second phase. For test–retest analyses, we included the 41 children with the same respondent (mother, father or legal representative) in both occasions. The median age of the children was 22,5 months at the first phase and 23,7 months at the second phase, with a predominance of girls (F:M = 1:0,6). The median time between the fillings of both questionnaires was 26 days. Globally, agreement values were good to excellent, similarly to the original work. Conclusion: In the present study we translated the Questionnaire on respiratory symptoms in preschool children and assessed its repeatability. Overall, we expect it to be a valuable tool for epidemiological studies and community-based screening programs, thus contributing to improve the management of infantile asthma / recurrent wheezing nationwide.

Automatic syllabification for danish text-to-speech systems

In this paper, a rule-based automatic syllabifier for Danish is described using the Maximal Onset Principle. Prior success rates of rule-based methods applied to Portuguese and Catalan syllabification modules were on the basis of this work. The system was implemented and tested using a very small set of rules. The results gave rise to 96.9% and 98.7% of word accuracy rate, contrary to our initial expectations, being Danish a language with a complex syllabic structure and thus difficult to be rule-driven. Comparison with data-driven syllabification system using artificial neural networks showed a higher accuracy rate of the former system.

An automatic voice pleasantness classification system based on prosodic and acoustic patterns of voice preference

In the last few years the number of systems and devices that use voice based interaction has grown significantly. For a continued use of these systems the interface must be reliable and pleasant in order to provide an optimal user experience. However there are currently very few studies that try to evaluate how good is a voice when the application is a speech based interface. In this paper we present a new automatic voice pleasantness classification system based on prosodic and acoustic patterns of voice preference. Our study is based on a multi-language database composed by female voices. In the objective performance evaluation the system achieved a 7.3% error rate.

Adaptação Cultural e Validação do Questionário: “Pregnancy Physical Activity Questionnaire” para a População Portuguesa

Introdução: A prática de atividade física tem vindo a ser incentivada por vários organismos nacionais e internacionais, visando a promoção de saúde. No entanto, não existe consenso em relação à influência e à intensidade ótima da atividade física durante a gravidez, tal facto pode dever-se à falta de instrumentos válidos que permitam a comparação entre os vários países do nível de atividade física das gestantes. Não foi identificado qualquer questionário válido que avaliasse a atividade física, durante a gravidez, para a população portuguesa. Objetivo: O objetivo deste estudo foi traduzir e adaptar o questionário PPAQ para a língua portuguesa e testar a fiabilidade e validade do mesmo. Métodos: A equivalência linguística foi obtida através da tradução e retrotradução e após a realização de um teste piloto a 6 mulheres grávidas. A fiabilidade foi obtida através do testereteste, com um intervalo de 7 dias, e analisada através do ICC, SEM e do método Bland Altman. A validade de critério foi testada, através dos resultados do PPAQ e do acelerómetro, pelo coeficiente de correlação de Pearson. Resultados: Os valores do ICC relativos à fiabilidade foram para o score total do PPAQ de 0,77, para as atividades sedentárias 0,87, para as atividade de baixa intensidade 0,76, para as atividades de intensidade moderada 0,76 e para as atividades de intensidade vigorosa 0,70. Na avaliação da validade obtivemos um valor de correlação de Pearson para atividade total de -0,030. Obtivemos correlação significante para as atividades de baixa intensidade (r=0,149; p=0,025), para as atividades de baixa intensidade em grávidas no 2º trimestre (r=0,256; p=0,010) e para as atividades de baixa intensidade e atividades vigorosas em mulheres no 3º trimestre gestacional (r=0,447; p=0,037 e r= 0,578; p=0,005, respetivamente). Conclusão: O PPAQ apresenta uma fiabilidade excelente e validade semelhante à versão original.

Automatic Small Bowel Tumor Diagnosis by Using Multi-Scale Wavelet-Based Analysis in Wireless Capsule Endoscopy Images

BACKGROUND: Wireless capsule endoscopy has been introduced as an innovative, non-invasive diagnostic technique for evaluation of the gastrointestinal tract, reaching places where conventional endoscopy is unable to. However, the output of this technique is an 8 hours video, whose analysis by the expert physician is very time consuming. Thus, a computer assisted diagnosis tool to help the physicians to evaluate CE exams faster and more accurately is an important technical challenge and an excellent economical opportunity. METHOD: The set of features proposed in this paper to code textural information is based on statistical modeling of second order textural measures extracted from co-occurrence matrices. To cope with both joint and marginal non-Gaussianity of second order textural measures, higher order moments are used. These statistical moments are taken from the two-dimensional color-scale feature space, where two different scales are considered. Second and higher order moments of textural measures are computed from the co-occurrence matrices computed from images synthesized by the inverse wavelet transform of the wavelet transform containing only the selected scales for the three color channels. The dimensionality of the data is reduced by using Principal Component Analysis. RESULTS: The proposed textural features are then used as the input of a classifier based on artificial neural networks. Classification performances of 93.1% specificity and 93.9% sensitivity are achieved on real data. These promising results open the path towards a deeper study regarding the applicability of this algorithm in computer aided diagnosis systems to assist physicians in their clinical practice.

Customer Lifetime Value: a framework for application in the insurance industry - Building a Business Process to Generate and Maintain an Automatic Estimation Agent

Research Project submited as partial fulfilment for the Master Degree in Statistics and Information Management

Automatic information retrieval through text-mining

The dissertation presented for obtaining the Master’s Degree in Electrical Engineering and Computer Science, at Universidade Nova de Lisboa, Faculdade de Ciências e Tecnologia

A semi-automatic approach to code smells detection

Eradication of code smells is often pointed out as a way to improve readability, extensibility and design in existing software. However, code smell detection remains time consuming and error-prone, partly due to the inherent subjectivity of the detection processes presently available. In view of mitigating the subjectivity problem, this dissertation presents a tool that automates a technique for the detection and assessment of code smells in Java source code, developed as an Eclipse plugin. The technique is based upon a Binary Logistic Regression model that uses complexity metrics as independent variables and is calibrated by expert‟s knowledge. An overview of the technique is provided, the tool is described and validated by an example case study.

Risky behavior regarding drug use and HIV infection: an Internet questionnaire coupled with short education texts for Portuguese speakers

Surveys of risky behavior relating to HIV/AIDS are generally made for groups at risk of infection, for which HIV/AIDS prevalence is usually expected to be higher than in the general population. Therefore, an educational homepage in Portuguese was created on the Internet to inform/ask internauts regarding knowledge and behavior. The internauts were classified as adolescents (13 to 25 years) and adults (>25 years). The number of STDs was reported as 1. 8 ± 2. 6 infections (range: 1 to 20 infections); 43% used condoms during sexual intercourse. Alcohol consumption was reported by 63% and illicit drug use by 32% (marijuana 24% and inhalants 15%). Among the adolescents, 31% did not classified alcohol as a drug. The adults more frequently reported homosexuality, anal intercourse and STDs, although the adolescents also presented high rates of risky behavior. These results show the need to reach out to internauts through better control strategies. Different types of strategies must be encouraged, in order to reach people that use this means of communication and entertainment.

Automatic delimitation of the clinical region of interest in ultra-wide field of view images of the retina

Retinal ultra-wide field of view images (fundus images) provides the visu-alization of a large part of the retina though, artifacts may appear in those images. Eyelashes and eyelids often cover the clinical region of interest and worse, eye-lashes can be mistaken with arteries and/or veins when those images are put through automatic diagnosis or segmentation software creating, in those cases, the appearance of false positives results. Correcting this problem, the first step in the development of qualified auto-matic diseases diagnosis programs can be done and in that way the development of an objective tool to assess diseases eradicating the human error from those processes can also be achieved. In this work the development of a tool that automatically delimitates the clinical region of interest is proposed by retrieving features from the images that will be analyzed by an automatic classifier. This automatic classifier will evaluate the information and will decide which part of the image is of interest and which part contains artifacts. The results were validated by implementing a software in C# language and validated through a statistical analysis. From those results it was confirmed that the methodology presented is capable of detecting artifacts and selecting the clin-ical region of interest in fundus images of the retina.

Automatic extraction of concepts from texts and applications

The extraction of relevant terms from texts is an extensively researched task in Text- Mining. Relevant terms have been applied in areas such as Information Retrieval or document clustering and classification. However, relevance has a rather fuzzy nature since the classification of some terms as relevant or not relevant is not consensual. For instance, while words such as "president" and "republic" are generally considered relevant by human evaluators, and words like "the" and "or" are not, terms such as "read" and "finish" gather no consensus about their semantic and informativeness. Concepts, on the other hand, have a less fuzzy nature. Therefore, instead of deciding on the relevance of a term during the extraction phase, as most extractors do, I propose to first extract, from texts, what I have called generic concepts (all concepts) and postpone the decision about relevance for downstream applications, accordingly to their needs. For instance, a keyword extractor may assume that the most relevant keywords are the most frequent concepts on the documents. Moreover, most statistical extractors are incapable of extracting single-word and multi-word expressions using the same methodology. These factors led to the development of the ConceptExtractor, a statistical and language-independent methodology which is explained in Part I of this thesis. In Part II, I will show that the automatic extraction of concepts has great applicability. For instance, for the extraction of keywords from documents, using the Tf-Idf metric only on concepts yields better results than using Tf-Idf without concepts, specially for multi-words. In addition, since concepts can be semantically related to other concepts, this allows us to build implicit document descriptors. These applications led to published work. Finally, I will present some work that, although not published yet, is briefly discussed in this document.

What makes people stay in or leave shrinking cities? An empirical study from Portugal

The relative attractiveness of cities as places to live determines population movements in or out of them. Understanding the appealing features of a city is fundamental to local governments, particularly for cities facing population decline. Pull and push attributes of cities can include economic aspects, the availability of amenities and psychological constructs, initiating a discussion around which factors are more relevant in explaining migration. However, a pull–push approach has been underexplored in studies of shrinking cities. In the present study, we contribute to the discussion by identifying pull and push factors in Portuguese shrinking cities. Data were collected using a face-to-face questionnaire survey of 701 residents in four shrinking cities: Oporto, Barreiro, Peso da Régua and Moura. Factor analysis and automatic linear modelling were used to analyse the data. Our results support previous findings that the economic activity of a city is the most relevant feature for retaining residents. However, other characteristics specific to each city, especially those related to heritage and natural beauty, are also shown to influence a city’s attractiveness as a place to live. The cause of population shrinkage is also found to influence residents’ assessments of the pull and push attributes of each city. Furthermore, the results show the relevance of social ties and of place attachment to inhabitants’ intention to continue living in their city of residence.

Detection of purkinje images for automatic positioning of fixation target and interferometric measurements of anterior eye chamber

In cataract surgery, the eye’s natural lens is removed because it has gone opaque and doesn’t allow clear vision any longer. To maintain the eye’s optical power, a new artificial lens must be inserted. Called Intraocular Lens (IOL), it needs to be modelled in order to have the correct refractive power to substitute the natural lens. Calculating the refractive power of this substitution lens requires precise anterior eye chamber measurements. An interferometry equipment, the AC Master from Zeiss Meditec, AG, was in use for half a year to perform these measurements. A Low Coherence Interferometry (LCI) measurement beam is aligned with the eye’s optical axis, for precise measurements of anterior eye chamber distances. The eye follows a fixation target in order to make the visual axis align with the optical axis. Performance problems occurred, however, at this step. Therefore, there was a necessity to develop a new procedure that ensures better alignment between the eye’s visual and optical axes, allowing a more user friendly and versatile procedure, and eventually automatizing the whole process. With this instrument, the alignment between the eye’s optical and visual axes is detected when Purkinje reflections I and III are overlapped, as the eye follows a fixation target. In this project, image analysis is used to detect these Purkinje reflections’ positions, eventually automatically detecting when they overlap. Automatic detection of the third Purkinje reflection of an eye following a fixation target is possible with some restrictions. Each pair of detected third Purkinje reflections is used in automatically calculating an acceptable starting position for the fixation target, required for precise measurements of anterior eye chamber distances.