Analysis of Phenotypic Variation in Childhood Wheezing Disorders

Recurrent wheezing or asthma is a common problem in children that has increased considerably in prevalence in the past few decades. The causes and underlying mechanisms are poorly understood and it is thought that a numb er of distinct diseases causing similar symptoms are involved. Due to the lack of a biologically founded classification system, children are classified according to their observed disease related features (symptoms, signs, measurements) into phenotypes. The objectives of this PhD project were a) to develop tools for analysing phenotypic variation of a disease, and b) to examine phenotypic variability of wheezing among children by applying these tools to existing epidemiological data. A combination of graphical methods (multivariate co rrespondence analysis) and statistical models (latent variables models) was used. In a first phase, a model for discrete variability (latent class model) was applied to data on symptoms and measurements from an epidemiological study to identify distinct phenotypes of wheezing. In a second phase, the modelling framework was expanded to include continuous variability (e.g. along a severity gradient) and combinations of discrete and continuo us variability (factor models and factor mixture models). The third phase focused on validating the methods using simulation studies. The main body of this thesis consists of 5 articles (3 published, 1 submitted and 1 to be submitted) including applications, methodological contributions and a review. The main findings and contributions were: 1) The application of a latent class model to epidemiological data (symptoms and physiological measurements) yielded plausible pheno types of wheezing with distinguishing characteristics that have previously been used as phenotype defining characteristics. 2) A method was proposed for including responses to conditional questions (e.g. questions on severity or triggers of wheezing are asked only to children with wheeze) in multivariate modelling.ii 3) A panel of clinicians was set up to agree on a plausible model for wheezing diseases. The model can be used to generate datasets for testing the modelling approach. 4) A critical review of methods for defining and validating phenotypes of wheeze in children was conducted. 5) The simulation studies showed that a parsimonious parameterisation of the models is required to identify the true underlying structure of the data. The developed approach can deal with some challenges of real-life cohort data such as variables of mixed mode (continuous and categorical), missing data and conditional questions. If carefully applied, the approach can be used to identify whether the underlying phenotypic variation is discrete (classes), continuous (factors) or a combination of these. These methods could help improve precision of research into causes and mechanisms and contribute to the development of a new classification of wheezing disorders in children and other diseases which are difficult to classify.

The Impact of Pre-Operative Malperfusion on Outcome in Acute Type A Aortic Dissection: Results From the GERAADA Registry

BACKGROUND Malperfusion adversely affects outcomes in patients with acute type A aortic dissection, but reliable quantitative data are lacking. OBJECTIVES The aim of this study was to analyze the impact of various forms of malperfusion on early outcome. METHODS A total of 2,137 consecutive patients enrolled in GERAADA (German Registry for Acute Aortic Dissection Type A) who underwent surgery between 2006 and 2010, of whom 717 (33.6%) had any kind of pre-operative malperfusion, were retrospectively analyzed. RESULTS All-cause 30-day mortality was 16.9% and varied substantially according to the number of organ systems affected by malperfusion (none, 12.6%; 1 system, 21.3%; 2 systems, 30.9%; 3 systems, 43.4%; p < 0.001). Pre-operative cerebral malperfusion, comatose state, peripheral malperfusion, visceral malperfusion, involvement of supra-aortic branches, coronary malperfusion, and renal malperfusion were all independent predictors of developing any post-operative malperfusion syndrome. When survival was considered, age, peripheral malperfusion, involvement of supra-aortic branches, coronary malperfusion, spinal malperfusion, a primary entry in the descending aorta, and pre-operative comatose state were independent predictors, again with increasing significance. CONCLUSIONS Malperfusion remains a severe clinical condition with strong potential for adverse outcomes in patients undergoing surgery for acute type A aortic dissection. The GERAADA registry suggests that the impact of the number of organs involved and the type of malperfusion on outcome differs substantially. Introducing an appropriate classification system, such as "complicated" and uncomplicated" acute type A aortic dissection, might help predict individual risk as well as select a surgical strategy that may quickly resolve malperfusion.

Towards mapping the late Quaternary vegetation change of Europe

The number of well-dated pollen diagrams in Europe has increased considerably over the last 30 years and many of them have been submitted to the European Pollen Database (EPD). This allows for the construction of increasingly precise maps of Holocene vegetation change across the continent. Chronological information in the EPD has been expressed in uncalibrated radiocarbon years, and most chronologies to date are based on this time scale. Here we present new chronologies for most of the datasets stored in the EPD based on calibrated radiocarbon years. Age information associated with pollen diagrams is often derived from the pollen stratigraphy itself or from other sedimentological information. We reviewed these chronological tie points and assigned uncertainties to them. The steps taken to generate the new chronologies are described and the rationale for a new classification system for age uncertainties is introduced. The resulting chronologies are fit for most continental-scale questions. They may not provide the best age model for particular sites, but may be viewed as general purpose chronologies. Taxonomic particularities of the data stored in the EPD are explained. An example is given of how the database can be queried to select samples with appropriate age control as well as the suitable taxonomic level to answer a specific research question.

Hysteroscopic outpatient metroplasty to expand dysmorphic uteri (HOME-DU technique): a pilot study.

The new classification system of uterine anomalies from the European Society of Human Reproduction and Embryology and the European Society for Gynaecological Endoscopy defines T-shaped and tubular-shaped infantilis uteri as 'dysmorphic'. Such malformations have been proven to be associated with poor reproductive performance. A prospective observational study was conducted with 30 infertile women with dysmorphic uterus who underwent the novel Hysteroscopic Outpatient Metroplasty to Expand Dysmorphic Uteri (HOME-DU ) technique. Incisions are made on the uterine walls with a 5 Fr bipolar electrode. The procedure was conducted in outpatients under conscious sedation, using a 5-mm office hysteroscope. The technique was successful in all cases without complications. A net increase of uterine volume was found, as measured at hysteroscopy and three-dimensional transvaginal ultrasound (P < 0.001). Uterine morphology improved in all patients but one. At mean follow-up of 15 months, clinical pregnancy rate was 57% and term delivery rate 65%. These early data support HOME-DU as safe and effective in expanding the volume and normalizing the appearance of the uterine cavity of dysmorphic uteri. Although the cohort was small, pregnancy and live births outcomes were favourable in this poor-prognosis group, implying desirable benefits, which should be compared with other techniques.

Percutaneous embolization of arteriovenous malformations at the plantar aspect of the foot

Peripheral arteriovenous malformations (AVM) remain most challenging among various congenital vascular malformations to be treated. Here we present three illustrative patients with Yakes type IIIb and type IV AVM at the plantar aspect of the foot who were successfully treated by minimally invasive embolization. The value of the Yakes AVM classification system to guide the therapeutic decision making by directing specific therapeutic procedures to specific AVM types defined by their angioarchitecture is demonstrated. Direct percutaneous AVM puncture with coiling of aneurysmal outflow vein and subsequent ethanol embolization is shown. Finally, the report illustrates that several AVM types can coexist.

Validated sample ages from 823 EPD sites

The number of well-dated pollen diagrams in Europe has increased considerably over the last 30 years and many of them have been submitted to the European Pollen Database (EPD). This allows for the construction of increasingly precise maps of Holocene vegetation change across the continent. Chronological information in the EPD has been expressed in uncalibrated radiocarbon years, and most chronologies to date are based on this time scale. Here we present new chronologies for most of the datasets stored in the EPD based on calibrated radiocarbon years. Age information associated with pollen diagrams is often derived from the pollen stratigraphy itself or from other sedimentological information. We reviewed these chronological tie points and assigned uncertainties to them. The steps taken to generate the new chronologies are described and the rationale for a new classification system for age uncertainties is introduced. The resulting chronologies are fit for most continental-scale questions. They may not provide the best age model for particular sites, but may be viewed as general purpose chronologies. Taxonomic particularities of the data stored in the EPD are explained. An example is given of how the database can be queried to select samples with appropriate age control as well as the suitable taxonomic level to answer a specific research question.

Geotechnical engineering characterization of sediments at DSDP Leg 72 Holes

The hydraulic piston coring device (HPC-15) allows recovery of deep ocean sediments with minimal disturbance. The device was used during Leg 72 of the Deep Sea Drilling Project (DSDP) aboard the Glomar Challenger. Core samples were recovered from bore holes in the Rio Grande Rise in the southwest Atlantic Ocean. Relatively undisturbed sediment cores were obtained from Holes 515A, 516, 517, and 518. The results of shipboard physical property measurements and on-shore geotechnical laboratory tests on these cores are presented in this chapter. A limited number of 0.3 m cores were obtained and used in a series of geotechnical tests, including one-dimensional consolidation, direct shear, Atterburg limit, particle size analysis, and specific gravity tests. Throughout the testing program, attention was focused on assessment of sample disturbance associated with the HPC-15 coring device. The HPC-15 device limits sample disturbance reasonably well in terrigenous muds (clays). However, sample disturbance associated with coring calcareous sediments (nannofossil-foraminifer oozes) is severe. The noncohesive, granular behavior of the calcareous sediments is vulnerable to severe disturbance, because of the design of the sampling head on the device at the time of Leg 72. A number of modifications to the sampling head design are recommended and discussed in this chapter. The modifications will improve sample quality for testing purposes and provide longer unbroken core samples by reducing friction between the sediment column and the sampling tool.

A fractional vegetation cover remote sensing product on pan-arctic scale with link to geotiff image

The paper presents first results of a pan-boreal scale land cover harmonization and classification. A methodology is presented that combines global and regional vegetation datasets to extract percentage cover information for different vegetation physiognomy and barren for the pan-arctic region within the ESA Data User Element Permafrost. Based on the legend description of each land cover product the datasets are harmonized into four LCCS (Land Cover Classification System) classifiers which are linked to the MODIS Vegetation Continuous Field (VCF) product. Harmonized land cover and Vegetation Continuous Fields products are combined to derive a best estimate of percentage cover information for trees, shrubs, herbaceous and barren areas for Russia. Future work will concentrate on the expansion of the developed methodology to the pan-arctic scale. Since the vegetation builds an isolation layer, which protects the permafrost from heat and cold temperatures, a degradation of this layer due to fire strongly influences the frozen conditions in the soil. Fire is an important disturbance factor which affects vast processes and dynamics in ecosystems (e.g. biomass, biodiversity, hydrology, etc.). Especially in North Eurasia the fire occupancy has dramatically increased in the last 50 years and has doubled in the 1990s with respect to the last five decades. A comparison of global and regional fire products has shown discrepancies between the amounts of burn scars detected by different algorithms and satellite data.

At surface behaviour at location on spot of southern elephant seals from Marion Island in 2004 with links to datasets

A novel classification system was applied to the sea level anomaly (SLA) environment around Marion Island. We classified the SLA seascape into habitat types and calculated percentage of habitat use of ten juvenile southern elephant seals (SES). Movements were compared to SLA and SLA slope values indicative of ocean eddy features. This classification provides a measure of habitat change due to seasonal fluctuations in SLA. Some of the seals made two migrations in different seasons, each of similar duration and proportions of potential foraging behaviour. The seals in this study did not use any intense eddy features, but their behaviours varied with SLA class. Potential foraging behaviour was positively influenced by negative SLA values (i.e. areas of below average sea surface height). Searching behaviour during the winter was more likely at eddy edges where high SLA slope values correlated with low SLA values. Though the seals did not forage within newly spawned eddies, they did forage near the sub-Antarctic front. Plankton and other biological resources transported by eddies formed at the subtropical convergence zone are evidently concentrated in this region and enhance the food chain there, forming a foraging ground for juvenile SES from Marion Island.

(Table A1) Species list of the high Antarctic Weddell Sea food web

Human-induced habitat destruction, overexploitation, introduction of alien species and climate change are causing species to go extinct at unprecedented rates, from local to global scales. There are growing concerns that these kinds of disturbances alter important functions of ecosystems. Our current understanding is that key parameters of a community (e.g. its functional diversity, species composition, and presence/absence of vulnerable species) reflect an ecological network's ability to resist or rebound from change in response to pressures and disturbances, such as species loss. If the food web structure is relatively simple, we can analyse the roles of different species interactions in determining how environmental impacts translate into species loss. However, when ecosystems harbour species-rich communities, as is the case in most natural systems, then the complex network of ecological interactions makes it a far more challenging task to perceive how species' functional roles influence the consequences of species loss. One approach to deal with such complexity is to focus on the functional traits of species in order to identify their respective roles: for instance, large species seem to be more susceptible to extinction than smaller species. Here, we introduce and analyse the marine food web from the high Antarctic Weddell Sea Shelf to illustrate the role of species traits in relation to network robustness of this complex food web. Our approach was threefold: firstly, we applied a new classification system to all species, grouping them by traits other than body size; secondly, we tested the relationship between body size and food web parameters within and across these groups and finally, we calculated food web robustness. We addressed questions regarding (i) patterns of species functional/trophic roles, (ii) relationships between species functional roles and body size and (iii) the role of species body size in terms of network robustness. Our results show that when analyzing relationships between trophic structure, body size and network structure, the diversity of predatory species types needs to be considered in future studies.

Global Fuelbed Dataset

This dataset presents the first global fuel map, containing all the parameters required to be input in the Fuel Characteristic Classification System (FCCS). The dataset was developed from different spatial variables, both based on satellite Earth observation products and fuel databases, and is comprised by a global fuelbed map and a database that includes the parameters of each fuelbed that affect fire behavior and effects. A total of 274 fuelbeds were created and parameterized, and can be input into FCCS to obtain fire potentials, surface fire behavior and carbon biomass for each fuelbed. The global fuel dataset can be used for a varied range of applications, including fire danger assessment, fire behavior estimations, fuel consumption calculations and emissions inventories.

A Method for Developing Ontologies from User-Generated Classication Systems

Abstract Web 2.0 applications enabled users to classify information resources using their own vocabularies. The bottom-up nature of these user-generated classification systems have turned them into interesting knowledge sources, since they provide a rich terminology generated by potentially large user communities. Previous research has shown that it is possible to elicit some emergent semantics from the aggregation of individual classifications in these systems. However the generation of ontologies from them is still an open research problem. In this thesis we address the problem of how to tap into user-generated classification systems for building domain ontologies. Our objective is to design a method to develop domain ontologies from user-generated classifications systems. To do so, we rely on ontologies in the Web of Data to formalize the semantics of the knowledge collected from the classification system. Current ontology development methodologies have recognized the importance of reusing knowledge from existing resources. Thus, our work is framed within the NeOn methodology scenario for building ontologies by reusing and reengineering non-ontological resources. The main contributions of this work are: An integrated method to develop ontologies from user-generated classification systems. With this method we extract a domain terminology from the classification system and then we formalize the semantics of this terminology by reusing ontologies in the Web of Data. Identification and adaptation of existing techniques for implementing the activities in the method so that they can fulfill the requirements of each activity. A novel study about emerging semantics in user-generated lists. Resumen La web 2.0 permitió a los usuarios clasificar recursos de información usando su propio vocabulario. Estos sistemas de clasificación generados por usuarios son recursos interesantes para la extracción de conocimiento debido principalmente a que proveen una extensa terminología generada por grandes comunidades de usuarios. Se ha demostrado en investigaciones previas que es posible obtener una semántica emergente de estos sistemas. Sin embargo la generación de ontologías a partir de ellos es todavía un problema de investigación abierto. Esta tesis trata el problema de cómo aprovechar los sistemas de clasificación generados por usuarios en la construcción de ontologías de dominio. Así el objetivo de la tesis es diseñar un método para desarrollar ontologías de dominio a partir de sistemas de clasificación generados por usuarios. El método propuesto reutiliza conceptualizaciones existentes en ontologías publicadas en la Web de Datos para formalizar la semántica del conocimiento que se extrae del sistema de clasificación. Por tanto, este trabajo está enmarcado dentro del escenario para desarrollar ontologías mediante la reutilización y reingeniería de recursos no ontológicos que se ha definido en la Metodología NeOn. Las principales contribuciones de este trabajo son: Un método integrado para desarrollar una ontología de dominio a partir de sistemas de clasificación generados por usuarios. En este método se extrae una terminología de dominio del sistema de clasificación y posteriormente se formaliza su semántica reutilizando ontologías en la Web de Datos. La identificación y adaptación de un conjunto de técnicas para implementar las actividades propuestas en el método de tal manera que puedan cumplir automáticamente los requerimientos de cada actividad. Un novedoso estudio acerca de la semántica emergente en las listas generadas por usuarios en la Web.

Mining recurring concepts in a dynamic feature space

Most data stream classification techniques assume that the underlying feature space is static. However, in real-world applications the set of features and their relevance to the target concept may change over time. In addition, when the underlying concepts reappear, reusing previously learnt models can enhance the learning process in terms of accuracy and processing time at the expense of manageable memory consumption. In this paper, we propose mining recurring concepts in a dynamic feature space (MReC-DFS), a data stream classification system to address the challenges of learning recurring concepts in a dynamic feature space while simultaneously reducing the memory cost associated with storing past models. MReC-DFS is able to detect and adapt to concept changes using the performance of the learning process and contextual information. To handle recurring concepts, stored models are combined in a dynamically weighted ensemble. Incremental feature selection is performed to reduce the combined feature space. This contribution allows MReC-DFS to store only the features most relevant to the learnt concepts, which in turn increases the memory efficiency of the technique. In addition, an incremental feature selection method is proposed that dynamically determines the threshold between relevant and irrelevant features. Experimental results demonstrating the high accuracy of MReC-DFS compared with state-of-the-art techniques on a variety of real datasets are presented. The results also show the superior memory efficiency of MReC-DFS.

Evaluación de la contribución y el impacto de las tecnologías del habla en la detección automática del Síndrome de la Apnea Obstructiva del Sueño = Contributions and impact assessment of speech technologies on the automatic detection of severe Obstructive Sleep Apnea syndrome

La presente Tesis analiza las posibilidades que ofrecen en la actualidad las tecnologías del habla para la detección de patologías clínicas asociadas a la vía aérea superior. El estudio del habla que tradicionalmente cubre tanto la producción como el proceso de transformación del mensaje y las señales involucradas, desde el emisor hasta alcanzar al receptor, ofrece una vía de estudio alternativa para estas patologías. El hecho de que la señal emitida no solo contiene este mensaje, sino también información acerca del locutor, ha motivado el desarrollo de sistemas orientados a la identificación y verificación de la identidad de los locutores. Estos trabajos han recibido recientemente un nuevo impulso, orientándose tanto hacia la caracterización de rasgos que son comunes a varios locutores, como a las diferencias existentes entre grabaciones de un mismo locutor. Los primeros resultan especialmente relevantes para esta Tesis dado que estos rasgos podrían evidenciar la presencia de características relacionadas con una cierta condición común a varios locutores, independiente de su identidad. Tal es el caso que se enfrenta en esta Tesis, donde los rasgos identificados se relacionarían con una de la patología particular y directamente vinculada con el sistema de físico de conformación del habla. El caso del Síndrome de Apneas Hipopneas durante el Sueno (SAHS) resulta paradigmático. Se trata de una patología con una elevada prevalencia mundo, que aumenta con la edad. Los pacientes de esta patología experimentan episodios de cese involuntario de la respiración durante el sueño, que se prolongan durante varios segundos y que se reproducen a lo largo de la noche impidiendo el correcto descanso. En el caso de la apnea obstructiva, estos episodios se deben a la imposibilidad de mantener un camino abierto a través de la vía aérea, de forma que el flujo de aire se ve interrumpido. En la actualidad, el diagnostico de estos pacientes se realiza a través de un estudio polisomnográfico, que se centra en el análisis de los episodios de apnea durante el sueño, requiriendo que el paciente permanezca en el hospital durante una noche. La complejidad y el elevado coste de estos procedimientos, unidos a las crecientes listas de espera, han evidenciado la necesidad de contar con técnicas rápidas de detección, que si bien podrían no obtener tasas tan elevadas, permitirían reorganizar las listas de espera en función del grado de severidad de la patología en cada paciente. Entre otros, los sistemas de diagnostico por imagen, así como la caracterización antropométrica de los pacientes, han evidenciado la existencia de patrones anatómicos que tendrían influencia directa sobre el habla. Los trabajos dedicados al estudio del SAHS en lo relativo a como esta afecta al habla han sido escasos y algunos de ellos incluso contradictorios. Sin embargo, desde finales de la década de 1980 se conoce la existencia de patrones específicos relativos a la articulación, la fonación y la resonancia. Sin embargo, su descripción resultaba difícilmente aprovechable a través de un sistema de reconocimiento automático, pero apuntaba la existencia de un nexo entre voz y SAHS. En los últimos anos las técnicas de procesado automático han permitido el desarrollo de sistemas automáticos que ya son capaces de identificar diferencias significativas en el habla de los pacientes del SAHS, y que los distinguen de los locutores sanos. Por contra, poco se conoce acerca de la conexión entre estos nuevos resultados, los sé que habían obtenido en el pasado y la patogénesis del SAHS. Esta Tesis continua la labor desarrollada en este ámbito considerando específicamente: el estudio de la forma en que el SAHS afecta el habla de los pacientes, la mejora en las tasas de clasificación automática y la combinación de la información obtenida con los predictores utilizados por los especialistas clínicos en sus evaluaciones preliminares. Las dos primeras tareas plantean problemas simbióticos, pero diferentes. Mientras el estudio de la conexión entre el SAHS y el habla requiere de modelos acotados que puedan ser interpretados con facilidad, los sistemas de reconocimiento se sirven de un elevado número de dimensiones para la caracterización y posterior identificación de patrones. Así, la primera tarea debe permitirnos avanzar en la segunda, al igual que la incorporación de los predictores utilizados por los especialistas clínicos. La Tesis aborda el estudio tanto del habla continua como del habla sostenida, con el fin de aprovechar las sinergias y diferencias existentes entre ambas. En el análisis del habla continua se tomo como punto de partida un esquema que ya fue evaluado con anterioridad, y sobre el cual se ha tratado la evaluación y optimización de la representación del habla, así como la caracterización de los patrones específicos asociados al SAHS. Ello ha evidenciado la conexión entre el SAHS y los elementos fundamentales de la señal de voz: los formantes. Los resultados obtenidos demuestran que el éxito de estos sistemas se debe, fundamentalmente, a la capacidad de estas representaciones para describir dichas componentes, obviando las dimensiones ruidosas o con poca capacidad discriminativa. El esquema resultante ofrece una tasa de error por debajo del 18%, sirviéndose de clasificadores notablemente menos complejos que los descritos en el estado del arte y de una única grabación de voz de corta duración. En relación a la conexión entre el SAHS y los patrones observados, fue necesario considerar las diferencias inter- e intra-grupo, centrándonos en la articulación característica del locutor, sustituyendo los complejos modelos de clasificación por el estudio de los promedios espectrales. El resultado apunta con claridad hacia ciertas regiones del eje de frecuencias, sugiriendo la existencia de un estrechamiento sistemático en la sección del tracto en la región de la orofaringe, ya prevista en la patogénesis de este síndrome. En cuanto al habla sostenida, se han reproducido los estudios realizados sobre el habla continua en grabaciones de la vocal /a/ sostenida. Los resultados son cualitativamente análogos a los anteriores, si bien en este caso las tasas de clasificación resultan ser más bajas. Con el objetivo de identificar el sentido de este resultado se reprodujo el estudio de los promedios espectrales y de la variabilidad inter e intra-grupo. Ambos estudios mostraron importantes diferencias con los anteriores que podrían explicar estos resultados. Sin embargo, el habla sostenida ofrece otras oportunidades al establecer un entorno controlado para el estudio de la fonación, que también había sido identificada como una fuente de información para la detección del SAHS. De su estudio se pudo observar que, en el conjunto de datos disponibles, no existen variaciones que pudieran asociarse fácilmente con la fonación. Únicamente aquellas dimensiones que describen la distribución de energía a lo largo del eje de frecuencia evidenciaron diferencias significativas, apuntando, una vez más, en la dirección de las resonancias espectrales. Analizados los resultados anteriores, la Tesis afronta la fusión de ambas fuentes de información en un único sistema de clasificación. Con ello es posible mejorar las tasas de clasificación, bajo la hipótesis de que la información presente en el habla continua y el habla sostenida es fundamentalmente distinta. Esta tarea se realizo a través de un sencillo esquema de fusión que obtuvo un 88.6% de aciertos en clasificación (tasa de error del 11.4%), lo que representa una mejora significativa respecto al estado del arte. Finalmente, la combinación de este clasificador con los predictores utilizados por los especialistas clínicos ofreció una tasa del 91.3% (tasa de error de 8.7%), que se encuentra dentro del margen ofrecido por esquemas más costosos e intrusivos, y que a diferencia del propuesto, no pueden ser utilizados en la evaluación previa de los pacientes. Con todo, la Tesis ofrece una visión clara sobre la relación entre el SAHS y el habla, evidenciando el grado de madurez alcanzado por la tecnología del habla en la caracterización y detección del SAHS, poniendo de manifiesto que su uso para la evaluación de los pacientes ya sería posible, y dejando la puerta abierta a futuras investigaciones que continúen el trabajo aquí iniciado. ABSTRACT This Thesis explores the potential of speech technologies for the detection of clinical disorders connected to the upper airway. The study of speech traditionally covers both the production process and post processing of the signals involved, from the speaker up to the listener, offering an alternative path to study these pathologies. The fact that utterances embed not just the encoded message but also information about the speaker, has motivated the development of automatic systems oriented to the identification and verificaton the speaker’s identity. These have recently been boosted and reoriented either towards the characterization of traits that are common to several speakers, or to the differences between records of the same speaker collected under different conditions. The first are particularly relevant to this Thesis as these patterns could reveal the presence of features that are related to a common condition shared among different speakers, regardless of their identity. Such is the case faced in this Thesis, where the traits identified would relate to a particular pathology, directly connected to the speech production system. The Obstructive Sleep Apnea syndrome (OSA) is a paradigmatic case for analysis. It is a disorder with high prevalence among adults and affecting a larger number of them as they grow older. Patients suffering from this disorder experience episodes of involuntary cessation of breath during sleep that may last a few seconds and reproduce throughout the night, preventing proper rest. In the case of obstructive apnea, these episodes are related to the collapse of the pharynx, which interrupts the air flow. Currently, OSA diagnosis is done through a polysomnographic study, which focuses on the analysis of apnea episodes during sleep, requiring the patient to stay at the hospital for the whole night. The complexity and high cost of the procedures involved, combined with the waiting lists, have evidenced the need for screening techniques, which perhaps would not achieve outstanding performance rates but would allow clinicians to reorganize these lists ranking patients according to the severity of their condition. Among others, imaging diagnosis and anthropometric characterization of patients have evidenced the existence of anatomical patterns related to OSA that have direct influence on speech. Contributions devoted to the study of how this disorder affects scpeech are scarce and somehow contradictory. However, since the late 1980s the existence of specific patterns related to articulation, phonation and resonance is known. By that time these descriptions were virtually useless when coming to the development of an automatic system, but pointed out the existence of a link between speech and OSA. In recent years automatic processing techniques have evolved and are now able to identify significant differences in the speech of OSAS patients when compared to records from healthy subjects. Nevertheless, little is known about the connection between these new results with those published in the past and the pathogenesis of the OSA syndrome. This Thesis is aimed to progress beyond the previous research done in this area by addressing: the study of how OSA affects patients’ speech, the enhancement of automatic OSA classification based on speech analysis, and its integration with the information embedded in the predictors generally used by clinicians in preliminary patients’ examination. The first two tasks, though may appear symbiotic at first, are quite different. While studying the connection between speech and OSA requires simple narrow models that can be easily interpreted, classification requires larger models including a large number dimensions for the characterization and posterior identification of the observed patterns. Anyhow, it is clear that any progress made in the first task should allow us to improve our performance on the second one, and that the incorporation of the predictors used by clinicians shall contribute in this same direction. The Thesis considers both continuous and sustained speech analysis, to exploit the synergies and differences between them. On continuous speech analysis, a conventional speech processing scheme, designed and evaluated before this Thesis, was taken as a baseline. Over this initial system several alternative representations of the speech information were proposed, optimized and tested to select those more suitable for the characterization of OSA-specific patterns. Evidences were found on the existence of a connection between OSA and the fundamental constituents of the speech: the formants. Experimental results proved that the success of the proposed solution is well explained by the ability of speech representations to describe these specific OSA-related components, ignoring the noisy ones as well those presenting low discrimination capabilities. The resulting scheme obtained a 18% error rate, on a classification scheme significantly less complex than those described in the literature and operating on a single speech record. Regarding the connection between OSA and the observed patterns, it was necessary to consider inter-and intra-group differences for this analysis, and to focus on the articulation, replacing the complex classification models by the long-term average spectra. Results clearly point to certain regions on the frequency axis, suggesting the existence of a systematic narrowing in the vocal tract section at the oropharynx. This was already described in the pathogenesis of this syndrome. Regarding sustained speech, similar experiments as those conducted on continuous speech were reproduced on sustained phonations of vowel / a /. Results were qualitatively similar to the previous ones, though in this case perfomance rates were found to be noticeably lower. Trying to derive further knowledge from this result, experiments on the long-term average spectra and intraand inter-group variability ratios were also reproduced on sustained speech records. Results on both experiments showed significant differences from the previous ones obtained from continuous speech which could explain the differences observed on peformance. However, sustained speech also provided the opportunity to study phonation within the controlled framework it provides. This was also identified in the literature as a source of information for the detection of OSA. In this study it was found that, for the available dataset, no sistematic differences related to phonation could be found between the two groups of speakers. Only those dimensions which relate energy distribution along the frequency axis provided significant differences, pointing once again towards the direction of resonant components. Once classification schemes on both continuous and sustained speech were developed, the Thesis addressed their combination into a single classification system. Under the assumption that the information in continuous and sustained speech is fundamentally different, it should be possible to successfully merge the two of them. This was tested through a simple fusion scheme which obtained a 88.6% correct classification (11.4% error rate), which represents a significant improvement over the state of the art. Finally, the combination of this classifier with the variables used by clinicians obtained a 91.3% accuracy (8.7% error rate). This is within the range of alternative, but costly and intrusive schemes, which unlike the one proposed can not be used in the preliminary assessment of patients’ condition. In the end, this Thesis has shed new light on the underlying connection between OSA and speech, and evidenced the degree of maturity reached by speech technology on OSA characterization and detection, leaving the door open for future research which shall continue in the multiple directions that have been pointed out and left as future work.

Clasificación de imágenes dermatoscópicas

Este proyecto presenta un software para el análisis de imágenes dermatoscópicas correspondiente a lesiones melanocíticas, con el fin de clasificarlas entre lesiones benignas y melanoma. El sistema realiza una segmentación automática de la lesión y la procesa en varas etapas, extrayendo características de relevancia diagnóstica: asimetría, colores, irregularidad del borde, y la presencia de estructuras como redes pigmentadas atípicas o velo azul-blanquecino. Proporciona además una herramienta para el etiquetado manual de estructuras adicionales. La clasificación automática de las lesiones se realiza en base a los métodos de diagnóstico más comúnmente utilizados: las reglas ABCD, Menzies, 7-point checklist, CASH y CHAOS & CLUES. El sistema de clasificación se evalúa sobre una base de datos de imágenes dermatoscópicas, y se realiza una comparativa de los resultados obtenidos por cada método de diagnóstico. ABSTRACT. This project presents a software for the analysis of dermoscopic images of melanocytic lesions, and their classification into benign lesions and melanoma. The system performs automatic segmentation of the lesion and goes through several stages of extraction of certain characteristics relevant to the diagnosis, such as asymmetry, border irregularity, or presence of structures like atypical pigmented network or blue-whitish veil. Automatic classification of the lesions is accomplished by means of the most commonly used diagnostic methods, such as ABCD and Menzies's rules, the 7-point checklist, CASH, and CHAOS & CLUES. The classification system is evaluated by using a dermoscopic image database, and a comparison of the results yielded by the different diagnostic methods is performed.