Methods for determining clusters of data under- specified constraints have recently gained popularity. Although general constraints may be used, we focus on clustering methods with the constraint of a minimal cluster size. In this dissertation, we propose two constrained k-means algorithms: Linear Programming Algorithm (LPA) and Genetic Constrained K-means Algorithm (GCKA). Linear Programming Algorithm modifies the k-means algorithm into a linear programming problem with constraints requiring that each cluster have m or more subjects. In order to achieve an acceptable clustering solution, we run the algorithm with a large number of random sets of initial seeds, and choose the solution with minimal Root Mean Squared Error (RMSE) as our final solution for a given data set. We evaluate LPA with both generic data and simulated data and the results indicate that LPA can obtain a reasonable clustering solution. Genetic Constrained K-Means Algorithm (GCKA) hybridizes the Genetic Algorithm with a constrained k-means algorithm. We define Selection Operator, Mutation Operator and Constrained K-means operator. Using finite Markov chain theory, we prove that the GCKA converges in probability to the global optimum. We test the algorithm with several datasets. The analysis shows that we can achieve a good clustering solution by carefully choosing parameters such as population size, mutation probability and generation. We also propose a Bi-Nelder algorithm to search for an appropriate cluster number with minimal RMSE.

Beat tracking is a common research area within music information retrieval (MIR). Jazz is a musical genre that is commonly rich in rhythmical complexity, which makes beat tracking challenging. The aim of this study is to analyze how well beat tracking algorithms detect the beats of jazz instruments. In this study, drums and double-bass were recorded in an ensemble but on single tracks. Three modern beat tracking algorithms were examined: LibRosa Dynamic, Essentia Multi-Feature, and madmom RNN. The algorithms’ beat trackings were evaluated using common metrics: the F-measure, P-score and Cemgil accuracy. The results showed that bass tracks generally got consistent results from all algorithms. However, all algorithms struggled with octave errors (the detected number of beats is off by a factor of two) and off-beats. When music was played without restrictions to the beat rhythm, madmom RNN generally performed the best, which suggests that machine learning with RNN (recurrent neural networks) is a good approach for beat tracking on rhythmically complex tracks. / Beat tracking är ett vanligt område inom music information retrieval (MIR). Jazz är en musikgenre som ofta är rytmisk komplex, vilket kan göra beat tracking utmanande. Denna studies syfte är att analysera hur väl beat tracking-algoritmer detekterar taktslagen hos jazzinstrument. I denna studie spelades trummor och kontrabas in samtidigt, men på separata spår. Tre moderna algoritmer för beat tracking undersöktes: LibRosa Dynamic, Essentia Multi-Feature, samt madmom RNN. Algoritmernas taktslagsestimeringar utvärderades utifrån vanligen tillämpade mätetal: F-measure, P-score och Cemgil accuracy. Resultaten visade att basspåren generellt sett gav konsekventa resultat från alla algoritmer. Däremot fick alla algoritmer oktavfel (antalet detekterade beats är fel med en faktor två) och off-beats. När musiken framfördes utan restriktioner i spelade taktslag presterade madmom RNN generellt bäst. Detta pekar mot att maskininlärning med RNN (recurrent neural networks) är en bra metod för beat tracking av rytmiskt komplexa spår.

Constantemente, uma variedade de novos métodos de processamento de imagens é apresentada à comunidade. Porém poucos têm provado sua utilidade na rotina clínica. A análise e comparação de diferentes abordagens por meio de uma mesma metodologia são essenciais para a qualificação do projeto de um algoritmo. Porém, é difícil comparar o desempenho e adequabilidade de diferentes algoritmos de uma mesma maneira. A principal razão deve-se à dificuldade para avaliar exaustivamente um software, ou pelo menos, testá-lo num conjunto abrangente e diversificado de casos clínicos. Muitas áreas - como o desenvolvimento de software e treinamentos em Medicina - necessitam de um conjunto diverso e abrangente de dados sobre imagens e informações associadas. Tais conjuntos podem ser utilizados para desenvolver, testar e avaliar novos softwares clínicos, utilizando dados públicos. Este trabalho propõe o desenvolvimento de um ambiente de base de imagens médicas de diferentes modalidades para uso livre em diferentes propósitos. Este ambiente - implementado como uma arquitetura de base distribuída de imagens - armazena imagens médicas com informações de aquisição, laudos, algoritmos de processamento de imagens, gold standards e imagens pós-processadas. O ambiente também possui um modelo de revisão de documentos que garante a qualidade dos conjuntos de dados. Como exemplo da facilidade e praticidade de uso, são apresentadas as avaliações de duas categorias de métodos de processamento de imagens médicas: segmentação e compressão. Em adição, a utilização do ambiente em outras atividades, como no projeto do arquivo didático digital do HC-FMUSP, demonstra a robustez da arquitetura proposta e sua aplicação em diferentes propósitos. / Constantly, a variety of new image processing methods are presented to the community. However, few of them have proved to be useful when used in clinical routine. The task of analyzing and comparing different algorithms, methods and applications through a sound testing is an essential qualification of algorithm design. However, it is usually very difficult to compare the performance and adequacy of different algorithms in the same way. The main reason is due to the difficulty to assess exhaustively the software, or at least using a comprehensive and diverse number of clinical cases for comparison. Several areas such as software development, image processing and medical training need a diverse and comprehensive dataset of images and related information. Such datasets could be used to develop, test and evaluate new medical software, using public data. This work presents the development of a free, online, multipurpose and multimodality medical image database environment. The environment, implemented such as a distributed medical image database, stores medical images, reports, image processing softwares, gold standards and post-processed images. Also, this environment implements a peer review model which assures the quality of all datasets. As an example of feasibility and easyness of use, it is shown the evaluation in two categories of medical image processing methods: segmentation and compression. In addition, the use of the set of applications proposed in this work in other activities, such as the HC-FMUSP digital teaching file, shows the robustness of the proposed architecture and its applicability on different purposes.

Constantemente, uma variedade de novos métodos de processamento de imagens é apresentada à comunidade. Porém poucos têm provado sua utilidade na rotina clínica. A análise e comparação de diferentes abordagens por meio de uma mesma metodologia são essenciais para a qualificação do projeto de um algoritmo. Porém, é difícil comparar o desempenho e adequabilidade de diferentes algoritmos de uma mesma maneira. A principal razão deve-se à dificuldade para avaliar exaustivamente um software, ou pelo menos, testá-lo num conjunto abrangente e diversificado de casos clínicos. Muitas áreas - como o desenvolvimento de software e treinamentos em Medicina - necessitam de um conjunto diverso e abrangente de dados sobre imagens e informações associadas. Tais conjuntos podem ser utilizados para desenvolver, testar e avaliar novos softwares clínicos, utilizando dados públicos. Este trabalho propõe o desenvolvimento de um ambiente de base de imagens médicas de diferentes modalidades para uso livre em diferentes propósitos. Este ambiente - implementado como uma arquitetura de base distribuída de imagens - armazena imagens médicas com informações de aquisição, laudos, algoritmos de processamento de imagens, gold standards e imagens pós-processadas. O ambiente também possui um modelo de revisão de documentos que garante a qualidade dos conjuntos de dados. Como exemplo da facilidade e praticidade de uso, são apresentadas as avaliações de duas categorias de métodos de processamento de imagens médicas: segmentação e compressão. Em adição, a utilização do ambiente em outras atividades, como no projeto do arquivo didático digital do HC-FMUSP, demonstra a robustez da arquitetura proposta e sua aplicação em diferentes propósitos. / Constantly, a variety of new image processing methods are presented to the community. However, few of them have proved to be useful when used in clinical routine. The task of analyzing and comparing different algorithms, methods and applications through a sound testing is an essential qualification of algorithm design. However, it is usually very difficult to compare the performance and adequacy of different algorithms in the same way. The main reason is due to the difficulty to assess exhaustively the software, or at least using a comprehensive and diverse number of clinical cases for comparison. Several areas such as software development, image processing and medical training need a diverse and comprehensive dataset of images and related information. Such datasets could be used to develop, test and evaluate new medical software, using public data. This work presents the development of a free, online, multipurpose and multimodality medical image database environment. The environment, implemented such as a distributed medical image database, stores medical images, reports, image processing softwares, gold standards and post-processed images. Also, this environment implements a peer review model which assures the quality of all datasets. As an example of feasibility and easyness of use, it is shown the evaluation in two categories of medical image processing methods: segmentation and compression. In addition, the use of the set of applications proposed in this work in other activities, such as the HC-FMUSP digital teaching file, shows the robustness of the proposed architecture and its applicability on different purposes.

Background: The large-scale analysis of phenomic data (i.e., full phenotypic traits of an organism, such as shape, metabolic substrates, and growth conditions) in microbial bioinformatics has been hampered by the lack of tools to rapidly and accurately extract phenotypic data from existing legacy text in the field of microbiology. To quickly obtain knowledge on the distribution and evolution of microbial traits, an information extraction system needed to be developed to extract phenotypic characters from large numbers of taxonomic descriptions so they can be used as input to existing phylogenetic analysis software packages. Results: We report the development and evaluation of Microbial Phenomics Information Extractor (MicroPIE, version 0.1.0). MicroPIE is a natural language processing application that uses a robust supervised classification algorithm (Support Vector Machine) to identify characters from sentences in prokaryotic taxonomic descriptions, followed by a combination of algorithms applying linguistic rules with groups of known terms to extract characters as well as character states. The input to MicroPIE is a set of taxonomic descriptions (clean text). The output is a taxon-by-character matrix-with taxa in the rows and a set of 42 pre-defined characters (e.g., optimum growth temperature) in the columns. The performance of MicroPIE was evaluated against a gold standard matrix and another student-made matrix. Results show that, compared to the gold standard, MicroPIE extracted 21 characters (50%) with a Relaxed F1 score > 0.80 and 16 characters (38%) with Relaxed F1 scores ranging between 0.50 and 0.80. Inclusion of a character prediction component (SVM) improved the overall performance of MicroPIE, notably the precision. Evaluated against the same gold standard, MicroPIE performed significantly better than the undergraduate students. Conclusion: MicroPIE is a promising new tool for the rapid and efficient extraction of phenotypic character information from prokaryotic taxonomic descriptions. However, further development, including incorporation of ontologies, will be necessary to improve the performance of the extraction for some character types.

Le transport de sédiments en rivières et torrents reste un phénomène mal compris en raison de la polydispersité des particules et de la ségrégation résultante. Il a été mené une étude expérimentale sur un canal permettant d’étudier la ségrégation en charriage d’un mélange de deux classes de billes. Le déplacement collectif des billes est enregistré sous la forme de séquences vidéos. Cette thèse traite des méthodes de traitement d’images développées pour analyser les données obtenues. Premièrement, nous avons développé une méthode de segmentation d’images pour étudier l’influence de l’infiltration de particules fines sur l’évolution d’un lit mobile. Avec cette méthode d’analyse, une étude expérimentale a permis de montrer que l’évolution de la pente du lit présente une décroissance exponentielle. Deuxièmement, nous avons optimisé les algorithmes déterministes de suivi de particules pour permettre l’étude des trajectoires sur l’intégralité du phénomène de ségrégation, ce qui n’était pas possible dans les travaux précédemment effectués à Irstea. Nous avons de plus mis en place des mesures d’évaluation et conçu des vérités terrains afin d’apprécier la qualité des résultats. Des gains de temps, cohérence, précision et mémoire ont été quantifiés. Troisièmement, nous avons développé un nouvel algorithme basé sur le filtrage particulaire à modèles multiples pour mieux gérer les dynamiques complexes des particules et gagner en robustesse. Cette approche permet de prendre en compte les erreurs du détecteur, les corriger et ainsi éviter des difficultés lors du suivi de trajectoires que nous rencontrons notamment avec l’algorithme déterministe / Sediment transport in rivers and mountain streams remains poorly understood partly due to the polydispersity of particles and resulting segregation. Experiments in a channel were carried out to study bedload transport of bimodal bead mixtures. The behavior of the beads is recorded through video sequences. This work is about the development of image processing methods to analyse the obtained data. Firstly, we developed a method of image segmentation to study the infiltration of fine particles and its influence on the evolution of bed mobility. Thanks to this method, an experimental study shows that the bed slope evolution follows an exponential decay. Secondly, we optimised deterministic tracking algorithms to enable the study of trajectories on long-duration phenomena of segregation, which was not possible with previous work done at Irstea. Moreover we set up relevant evaluation measures and elaborated ground truth sequences to quantify the results. We observed benefits in execution time, consistency, precision and memory. Thirdly, we developed a new algorithm based on multiple model particle filtering to better deal with complex dynamics of particles and to gain robustness. This approach allows taking unreliable detections into account, correcting them and thus avoiding difficulties in the target tracking as encountered with the deterministic algorithm

Driving is one of the most common and preferred forms of transport used in our actual society. However, according to studies, it is also one of the most dangerous. One solution to increase safety on the road is applying technology to automate and prevent avoidable human errors. Nevertheless, despite the efforts to obtain reliable systems, we have yet to find a reliable and safe enough solution for solving autonomous driving. One of the reasons is that many drives are done in conditions far from the ideal, with variable lighting conditions and fast-paced, unpredictable environments. This project develops and evaluates an algorithm that takes the input of dynamic vision sensors (DVS) and runs on neuromorphic spiking neural networks (SNN) to obtain a robust road lane tracking system. We present quantitative and qualitative metrics that evaluate the performance of lane recognition in low light conditions against conventional algorithms. This project is motivated by the main advantages of neuromorphic vision sensors: recognizing a high dynamic range and allowing a high-speed image capture. Another improvement of this system is the computational speed and power efficiency that characterize neuromorphic hardware based on spiking neural networks. The results obtained show a similar accuracy of this new algorithm compared to previous implementations on conventional hardware platforms. Most importantly, it accomplishes the proposed task with lower latency and computing power requirements than previous algorithms. / Att köra bil är ett av de vanligaste och mest populära transportsätten i vårt samhälle. Enligt forskningen är det också ett av de farligaste. En lösning för att öka säkerheten på vägarna är att med teknikens hjälp automatisera bilkörningen och på så sätt förebygga misstag som beror på den mänskliga faktorn. Trots ansträngningarna för att få fram tillförlitliga system har man dock ännu inte hittat en tillräckligt tillförlitlig och säker lösning för självkörande bilar. En av orsakerna till det är att många körningar sker under förhållanden som är långt ifrån idealiska, med varierande ljusförhållanden och oförutsägbara miljöer i höga hastigheter. I det här projektet utvecklar och utvärderar vi en algoritm som tar emot indata från dynamiska synsensorer (Dynamic Vision Sensors, DVS) och kör datan på neuromorfiska pulserande neuronnät (Spiking Neural Networks, SNN) för att skapa ett robust system för att läsa av vägbanan. Vi presenterar en kvantitativ och kvalitativ utvärdering av hur väl systemet läser av körbanans linjer i svagt ljus, och jämför därefter resultaten med dem för tidigare algoritmer. Detta projekt motiveras av de viktigaste fördelarna med neuromorfiska synsensorer: brett dynamiskt omfång och hög bildtagningshastighet. En annan fördel hos detta system är den korta beräkningstiden och den energieffektivitet som kännetecknar neuromorfisk hårdvara baserad på pulserande neuronnät. De resultat som erhållits visar att den nya algoritmen har en liknande noggrannhet som tidigare algoritmer på traditionella hårdvaruplattformar. I jämförelse med den traditionella tekniken, utför algoritmen i den föreliggande studien sin uppgift med kortare latenstid och lägre krav på processorkraft. / La conducción es una de las formas de transporte más comunes y preferidas en la actualidad. Sin embargo, diferentes estudios muestran que también es una de las más peligrosas. Una solución para aumentar la seguridad en la carretera es aplicar la tecnología para automatizar y prevenir los evitables errores humanos. No obstante, a pesar de los esfuerzos por conseguir sistemas fiables, todavía no hemos encontrado una solución suficientemente fiable y segura para resolver este reto. Una de las razones es el entorno de la conducción, en situaciones que distan mucho de las ideales, con condiciones de iluminación variables y entornos rápidos e imprevisibles. Este proyecto desarrolla y evalúa un algoritmo que toma la entrada de sensores de visión dinámicos (DVS) y ejecuta su computación en redes neuronales neuromórficas (SNN) para obtener un sistema robusto de seguimiento de carriles en carretera. Presentamos métricas cuantitativas y cualitativas que evalúan el rendimiento del reconocimiento de carriles en condiciones de poca luz, frente a algoritmos convencionales. Este proyecto está motivado por la validación de las ventajas de los sensores de visión neuromórficos: el reconocimiento de un alto rango dinámico y la captura de imágenes de alta velocidad. Otra de las mejoras que se espera de este sistema es la velocidad de procesamiento y la eficiencia energética que caracterizan al hardware neuromórfico basado en redes neuronales de impulsos. Los resultados obtenidos muestran una precisión similar entre el nuevo algoritmo en comparación con implementaciones anteriores en plataformas convencionales. Y lo que es más importante, realiza la tarea propuesta con menor latencia y requisitos de potencia de cálculo.

Atrial fibrillation is a common heart arrhythmia which is characterized by a missing or irregular contraction of the atria. The disease is a risk factor for other more serious diseases and the total medical costs in society are extensive. Therefore it would be beneficial to improve and optimize the prevention and detection of the disease.   Pulse palpation and heart auscultation can facilitate the detection of atrial fibrillation clinically, but the diagnosis is generally confirmed by an ECG examination. Today there are several algorithms that detect atrial fibrillation by analysing an ECG. A common method is to study the heart rate variability (HRV) and by different types of statistical calculations find episodes of atrial fibrillation which deviates from normal sinus rhythm.   Two algorithms for detection of atrial fibrillation have been evaluated in Matlab. One is based on the coefficient of variation and the other uses a logistic regression model. Training and testing of the algorithms were done with data from the Physionet MIT database. Several steps of signal processing were used to remove different types of noise and artefacts before the data could be used.   When testing the algorithms, the CV algorithm performed with a sensitivity of 91,38%, a specificity of 93,93% and accuracy of 92,92%, and the results of the logistic regression algorithm was a sensitivity of 97,23%, specificity of 93,79% and accuracy of 95,39%. The logistic regression algorithm performed better and was chosen for implementation in Java, where it achieved a sensitivity of 97,31%, specificity of 93,47% and accuracy of 95,25%. / Förmaksflimmer är en vanlig hjärtrytmrubbning som kännetecknas av en avsaknad eller oregelbunden kontraktion av förmaken. Sjukdomen är en riskfaktor för andra allvarligare sjukdomar och de totala kostnaderna för samhället är betydande. Det skulle därför vara fördelaktigt att effektivisera och förbättra prevention samt diagnostisering av förmaksflimmer.   Kliniskt diagnostiseras förmaksflimmer med hjälp av till exempel pulspalpation och auskultation av hjärtat, men diagnosen brukar fastställas med en EKG-undersökning. Det finns idag flertalet algoritmer för att detektera arytmin genom att analysera ett EKG. En av de vanligaste metoderna är att undersöka variabiliteten av hjärtrytmen (HRV) och utföra olika sorters statistiska beräkningar som kan upptäcka episoder av förmaksflimmer som avviker från en normal sinusrytm.   I detta projekt har två metoder för att detektera förmaksflimmer utvärderats i Matlab, en baseras på beräkningar av variationskoefficienten och den andra använder sig av logistisk regression. EKG som kommer från databasen Physionet MIT används för att träna och testa modeller av algoritmerna. Innan EKG-signalen kan användas måste den behandlas för att ta bort olika typer av brus och artefakter.   Vid test av algoritmen med variationskoefficienten blev resultatet en sensitivitet på 91,38%, en specificitet på 93,93% och en noggrannhet på 92,92%. För logistisk regression blev sensitiviteten 97,23%, specificiteten 93,79% och noggrannheten 95,39%. Algoritmen med logistisk regression presterade bättre och valdes därför för att implementeras i Java, där uppnåddes en sensitivitet på 91,31%, en specificitet på 93,47% och en noggrannhet på 95,25%.