Dissertations / Theses on the topic 'Diagnostic informatics'

Šio tyrimo tikslas yra sukurti, realizuoti ir išbandyti metodą leidžiantį atlikti naujo riedėjimo guolio elementų neardoma diagnostiką ir įvertinti surinkimo kokybę. Norint įvertinti guolio kokybę jo neišardant, pasitelkti virpesių pagreičio signalo stebėsenos ir analizės metodai. Atlikti sekantys uždaviniai siekiant įgyvendinti tyrimo tikslą: • Apžvelgti pramonėje naudojami guolių kokybės įvertinimo metodai. • Sukurtas ir realizuotas metodas naujo guolio kokybei nustatyti. • Praktiškai patikrinta metodo teikiama nauda. • Išnagrinėta suteptų ir nesuteptų guolių įtaką kokybės nustatymui. Atliekant virpesių signalų tyrimą sukurtu ir realizuotu guolių elementų virpesių pagreičio vidutinės kvadratinės vertės metodu, galima nustatyti guolio surinkimo kokybę, jo elementų pažeidimus, neatliekant ardymo kas būna neišvengiama taikant kitus metodus. Metodo efektyvumas patvirtintas atliekant tiriamųjų guolių išsamius mechaninius tyrimus, kurių metu buvo išardyti tyrimuose naudoti riedėjimo guoliai ir nuodugniai ištirta jų riedėjimo elementų kokybė ir gauti rezultatai palyginti su realizuoto metodo teikiamais rezultatais. Realizuoti pagalbiniai metodai kaip virpesių pagreičio skirtumo radimas, vyraujančių dažnių paieška, kurie suteikia papildomos informacijos apie guolį ir/ar jo kokybę. Sukurtas ir realizuotas virpesių pagreičio signalo centravimo metodas, patikrinta jo teikiama praktinė nauda.
The purpose of this thesis is to create, implement and test a method which would allow to evaluate the quality of newly made bearing and its elements, without the need of disassembling the bearing. To evaluate the quality of the bearing without disassembling it, vibration signal monitoring and analysis methods were used. The following tasks were completed in order to accomplish the goal of the thesis: • Reviewed bearing quality evaluation and failure detection methods used in industry. • Created and implemented a method which could evaluate the quality of a bearing. • Experimentally tested the efficiency of the method. • Investigated the used oil influence on the quality detection. The quality of the bearing and its elements can be measured using bearing elements vibration acceleration root mean square method. This method does not require the disassembling of the bearing which might not be avoided when using other methods. The efficiency of the method was proven by performing extensive mechanical analysis on studied bearings, which required to disassemble the bearing and check in detail the quality of the bearing. The collected results were compared with the ones received from the implemented method. Some additional methods were created and implemented which would give more information about the bearing and its quality. The vibration acceleration signal centering method was implemented and its efficiency was checked experimentally.

Cutaneous malignancies represent a quarter of all new cancer diagnoses in the UK. The key to reducing the tumours’ associated mortality and morbidity is early diagnosis and treatment. Prompt diagnosis remains predominately a clinical skill, but relatively little investigation of the cognitive psychology underpinning expertise in this domain has been undertaken. This thesis aims to improve understanding of these processes and investigate how lesional diagnostic expertise might be enhanced. A large database of diagnostically tagged images was captured specifically for this project. A series of separate studies were undertaken to give insight into how lesional diagnosis occurs and how it can be improved. The studies highlighted that non-analytical pattern recognition (NAPR) is likely to predominate in distinguishing malignant and non-malignant skin lesions and that the widely-promoted rules advocating analytical pattern recognition (APR) are not effective for discriminating melanoma from benign pigmented lesions. The keystone to promoting the development of NAPR and thus diagnostic expertise would seem to be increasing a novice’s personal library of examples with relevant feedback. Studies demonstrated that current undergraduate exposure was variable but universally sparse, so simulation by way of diagnostically tagged images was developed which showed accuracy could be improved by increased exposure. This improvement occurred in both a content specific and dose responsive manner. These studies also highlighted that the learning curves for skin lesions are not uniform. Further studies demonstrated that the choice of images had implications on the development of diagnostic expertise; suggesting it was important that these images represent clinical practice rather than “classic” examples traditionally advocated for teaching purposes. In addition, studies highlighted the potential benefit of the 3D models developed during this project. Building on the idea that a personal catalogue of relevant referent images was crucial to enhanced diagnostic accuracy, prototype software was developed to exteriorise the experts’ library of examples; in the tests described novices utilising the software delivered superior accuracy than medical students on the completion of their undergraduate teaching. In summation, the work described shows that by utilising dermato-informatic approaches lesional diagnostic competence can be improved significantly.

Medical imaging informatics is one of the important research areas in radiology that studies how information available on medical images is retrieved, analyzed, and enhanced. Recent development in medical imaging informatics has resulted in improvement of diagnostic accuracy based on imaging examinations, as well as efficiency in clinical workflow. Computer aided diagnosis (CAD) and electronic patient record system (ePR) are both topics in medical imaging informatics that have matured from research concepts into commercially available computerized systems in clinical environment. The current challenges are to further broaden their scope of applications. In this thesis project, I developed a CAD system for interpreting PET/CT examinations and an ePR system for patient data integration in neurosurgery suites. Specifically, the CAD system in this project was designed to automatically diagnose nasopharyngeal carcinoma (NPC) on Positron emission tomography/computed tomography (PET/CT) examinations, which aimed to detect and classify both the primary NPC and its nodal metastasis. The regions of interests (ROIs) were segmented from the PET images and registered onto the CT in order to combine the imaging features from both modalities and the a priori anatomical knowledge of the suspicious lesion. These combined features were then classified by a support vector machine (SVM) to generate the final diagnosis result. The system was validated with 25 PET/CT examinations from 10 patients suffering from NPC, and the result produced by the system was compared to the gold standard of lesions manually contoured by experienced radiologists. The results confirmed that the system successfully distinguished all 53 genuine lesions from the mimickers due to normal physiological uptake and artifacts that also produced potentially confusing signals. The second part of the project involved development of an electronic patient record system (ePR) that integrated all the myriad of images and different types of clinical information before, during, and after neurosurgery operations, in order to enhance efficiency of work flow in this unique clinical environment. The system comprises of pre-, intra-, and post-operation modules which correspond to the different stages of the neurosurgery. The pre-op module was developed to store and categorize all images and data before the procedure to assist the surgeons in planning operation. The intra-op module integrates all the input signals, waveforms, images and videos that are produced by different imaging and physiological monitoring devices in the operation room during the surgery, and displays all the relevant information in a single large screen in real time to ease monitoring of the procedure. The post-op module helps surgeons to review all the data acquired from all the prior stages for follow-up of the treatment outcome. One-tumor case was utilized to test the pre-op module, and the signals and waveforms simulators were used to evaluate the performance of the intra-op module. In summary, two different medical informatics systems, a CAD and an ePR system were developed. Both showed promising results in laboratory tests. Future work would involve performance enhancement and feedback of the systems, and ultimately evaluation of these systems in the clinical environment.
published_or_final_version
Diagnostic Radiology
Master
Master of Philosophy

Le sepsis tardif, défini comme une infection systémique chez les nouveaux nés âgés de plus de 3 jours, survient chez environ 7% à 10% de tous les nouveau-nés et chez plus de 25% des nouveau-nés de très faible poids de naissance qui sont hospitalisés dans les unités de soins intensifs néonatals (USIN). Les apnées et bradycardies (AB) spontanées récurrentes et graves sont parmi les principaux indicateurs précoces cliniques de l'infection systémique chez les prématurés. L'objectif de cette thèse est de déterminer si la variabilité du rythme cardiaque (VRC), la respiration et l'analyse de leurs relations aident au diagnostic de l'infection chez les nouveaux nés prématurés par des moyens non invasifs en USIN. Par conséquent, on a effectué l'analyse Mono-Voie (MV) et Bi-Voies (BV) sur deux groupes sélectionnés de nouveau-nés prématurés: sepsis (S) vs. non-sepsis (NS). (1) Tout d'abord, on a étudié la série RR non seulement par des méthodes de distribution (moy, varn, skew, kurt, med, SpAs), par les méthodes linéaire: le domaine temporel (SD, RMSSD) et dans le domaine fréquentiel (p_VLF, p_LF, p_HF), mais aussi par les méthodes non–linéaires: la théorie du chaos (alphas, alphaF) et la théorie de l'information (AppEn, SamEn, PermEn, Regul). Pour chaque méthode, nous étudions trois tailles de fenêtre 1024/2048/4096, puis nous comparons ces méthodes afin de trouver les meilleures façons de distinguer S de NS. Les résultats montrent que les indices alphaS, alphaF et SamEn sont les paramètres optimaux pour séparer les deux populations. (2) Ensuite, la question du couplage fonctionnel entre la VRC et la respiration nasale est adressée. Des relations linéaires et non-linéaires ont été explorées. Les indices linéaires sont la corrélation (r²), l'indice de la fonction de cohérence (Cohere) et la corrélation temps-fréquence (r2t,f) , tandis que le coefficient de régression non-linéaire (h²) a été utilisé pour analyser des relations non-linéaires. Nous avons calculé les deux directions de couplage pendant l'évaluation de l'indice h2 de régression non-linéaire. Enfin, à partir de l'ensemble du processus d'analyse, il est évident que les trois indices (r2tf_rn_raw_0p2_0p4, h2_rn_raw et h2_nr_raw) sont des moyens complémentaires pour le diagnostic du sepsis de façon non-invasive chez ces patients fragiles. (3) Après, l'étude de faisabilité de la détection du sepsis en USIN est réalisée sur la base des paramètres retenus lors des études MV et BV. Nous avons montré que le test proposé, basé sur la fusion optimale des six indices ci-dessus, conduit à de bonnes performances statistiques. En conclusion, les mesures choisies lors de l'analyse des signaux en MV et BV ont une bonne répétabilité et permettent de mettre en place un test en vue du diagnostic non invasif et précoce du sepsis. Le test proposé peut être utilisé pour fournir une alarme fiable lors de la survenue d'un épisode d'AB tout en exploitant les systèmes de monitoring actuels en USIN
Late-onset sepsis, defined as a systemic infection in neonates older than 3 days, occurs in approximately 10% of all neonates and in more than 25% of very low birth weight infants who are hospitalized in Neonatal Intensive Care Units (NICU). Recurrent and severe spontaneous apneas and bradycardias (AB) is one of the major clinical early indicators of systemic infection in the premature infant. Various hematological and biochemical markers have been evaluated for this indication but they are invasive procedures that cannot be repeated several times. The objective of this Ph.D dissertation was to determine if heart rate variability (HRV), respiration and the analysis of their relationships help to the diagnosis of infection in premature infants via non-invasive ways in NICU. Therefore, we carried out Mono-Channel (MC) and Bi-Channel (BC) Analysis in two selected groups of premature infants: sepsis (S) vs. non-sepsis (NS). (1) Firstly, we studied the RR series not only by distribution methods (moy, varn, skew, kurt, med, SpAs), by linear methods: time domain (SD, RMSSD) and frequency domain (p_VLF, p_LF, p_HF), but also by non-linear methods: chaos theory (alphaS, alphaF) and information theory (AppEn, SamEn, PermEn, Regul). For each method, we attempt three sizes of window 1024/2048/4096, and then compare these methods in order to find the optimal ways to distinguish S from NS. The results show that alphaS, alphaF and SamEn are optimal parameters to recognize sepsis from the diagnosis of late neonatal infection in premature infants with unusual and recurrent AB. (2) The question about the functional coupling of HRV and nasal respiration is addressed. Linear and non-linear relationships have been explored. Linear indexes were correlation (r²), coherence function (Cohere) and time-frequency index (r2t,f), while a non-linear regression coefficient (h²) was used to analyze non-linear relationships. We calculated two directions during evaluate the index h2 of non-linear regression. Finally, from the entire analysis process, it is obvious that the three indexes (r2tf_rn_raw_0p2_0p4, h2_rn_raw and h2_nr_raw) were complementary ways to diagnosticate sepsis in a non-invasive way, in such delicate patients.(3) Furthermore, feasibility study is carried out on the candidate parameters selected from MC and BC respectively. We discovered that the proposed test based on optimal fusion of 6 features shows good performance with the largest Area Under Curves (AUC) and the least Probability of False Alarm (PFA). As a conclusion, we believe that the selected measures from MC and BC signal analysis have a good repeatability and accuracy to test for the diagnosis of sepsis via non-invasive NICU monitoring system, which can reliably confirm or refute the diagnosis of infection at an early stage

This thesis evaluates how a head mounted display (HMD) can be used to increase usability compared to existing computer programs that are used during maintenance work on vehicles. Problems identified during a case study in a vehicle workshop are first described. As an attempt to solve some of the identified problems a prototype application using a HMD was developed. The prototype application aids the user during troubleshooting of systems on the vehicle by leading the mechanic with textual information and augmented reality (AR). Assessment of the prototype application was done by comparing it to the existing computer program and measuring error rate and time to completion for a predefined task. Usability was also measured using the System Usability Scale. The assessment showed that HMDs can provide higher usability in terms of efficiency and satisfaction. Furthermore, the thesis describes and discusses other possibilities and limitations that usage of HMDs and AR can lead to that were identified both from theory and during implementation.

The purpose of this study is to investigate the use of decisions support systems (DSS) by critical care physicians and to address the following questions: Does the use of a decision support system during diagnosis reduce diagnostic error and how are decision support systems used by critical care physicians? There are no studies that address these research questions in a clinical setting. The information assessment method (IAM) was used to guide the development of the survey questions. Critical care physicians from the University of Oklahoma Health Sciences Center were surveyed. Chi squared test for independence was used to determine the relationship between DSS use and diagnostic error rates. There were three main findings of the study: (1) use of a DSS by a critical care physician can decrease diagnostic error by up to 60%; (2) 56% of critical care physicians are using a DSS during diagnosis to learn something new, confirm something they already knew, and/or to reassure themselves; and (3) the increased use of a DSS by critical care physicians can lead to a decrease in the belief of the ability of a DSS to reduce diagnostic error.

Proficient categorization of facial expressions is crucial for normal social interaction. Neurophysiological, behavioural, event-related potential, lesion and functional neuroimaging techniques can be used to investigate the underlying brain mechanisms supporting this seemingly effortless process, and the associated arrangement of bilateral networks. These brain areas exhibit consistent and replicable activation patterns, and can be broadly defined to include visual (occipital and temporal), limbic (amygdala) and prefrontal (orbitofrontal) regions. Together, these areas support early perceptual processing, the formation of detailed representations and subsequent recognition of expressive faces. Despite the critical role of facial expressions in social communication and extensive work in this area, it is still not known how the brain decodes nonverbal signals in terms of expression-specific features. For these reasons, this thesis investigates the role of these so-called diagnostic facial features at three significant stages in expression recognition; the spatiotemporal inputs to the visual system, the dynamic integration of features in higher visual (occipitotemporal) areas, and early sensitivity to features in V1. In Chapter 1, the basic emotion categories are presented, along with the brain regions that are activated by these expressions. In line with this, the current cognitive theory of face processing reviews functional and anatomical dissociations within the distributed neural “face network”. Chapter 1 also introduces the way in which we measure and use diagnostic information to derive brain sensitivity to specific facial features, and how this is a useful tool by which to understand spatial and temporal organisation of expression recognition in the brain. In relation to this, hierarchical, bottom-up neural processing is discussed along with high-level, top-down facilitatory mechanisms. Chapter 2 describes an eye-movement study that reveals inputs to the visual system via fixations reflect diagnostic information use. Inputs to the visual system dictate the information distributed to cognitive systems during the seamless and rapid categorization of expressive faces. How we perform eye-movements during this task informs how task-driven and stimulus-driven mechanisms interact to guide the extraction of information supporting recognition. We recorded eye movements of observers who categorized the six basic categories of facial expressions. We use a measure of task-relevant information (diagnosticity) to discuss oculomotor behaviour, with focus on two findings. Firstly, fixated regions reveal expression differences. Secondly, by examining fixation sequences, the intersection of fixations with diagnostic information increases in a sequence of fixations. This suggests a top-down drive to acquire task-relevant information, with different functional roles for first and final fixations. A combination of psychophysical studies of visual recognition together with the EEG (electroencephalogram) signal is used to infer the dynamics of feature extraction and use during the recognition of facial expressions in Chapter 3. The results reveal a process that integrates visual information over about 50 milliseconds prior to the face-sensitive N170 event-related potential, starting at the eye region, and proceeding gradually towards lower regions. The finding that informative features for recognition are not processed simultaneously but in an orderly progression over a short time period is instructive for understanding the processes involved in visual recognition, and in particular the integration of bottom-up and top-down processes. In Chapter 4 we use fMRI to investigate the task-dependent activation to diagnostic features in early visual areas, suggesting top-down mechanisms as V1 traditionally exhibits only simple response properties. Chapter 3 revealed that diagnostic features modulate the temporal dynamics of brain signals in higher visual areas. Within the hierarchical visual system however, it is not known if an early (V1/V2/V3) sensitivity to diagnostic information contributes to categorical facial judgements, conceivably driven by top-down signals triggered in visual processing. Using retinotopic mapping, we reveal task-dependent information extraction within the earliest cortical representation (V1) of two features known to be differentially necessary for face recognition tasks (eyes and mouth). This strategic encoding of face images is beyond typical V1 properties and suggests a top-down influence of task extending down to the earliest retinotopic stages of visual processing. The significance of these data is discussed in the context of the cortical face network and bidirectional processing in the visual system. The visual cognition of facial expression processing is concerned with the interactive processing of bottom-up sensory-driven information and top-down mechanisms to relate visual input to categorical judgements. The three experiments presented in this thesis are summarized in Chapter 5 in relation to how diagnostic features can be used to explore such processing in the human brain leading to proficient facial expression categorization.

Rodon is a diagnostic tool developed by Sörman. SAAB’s interest in Rodon regards the possibility to use the tool for development and operations of aircraft systems. The main goal of this thesis was to evaluate the capacity of Rodon and determine how SAAB can use the diagnostic tool during development and operations.

The tool uses model based diagnosis with artificial intelligence for fault isolation which is a powerful approach. If Rodon is introduced at SAAB, then detailed models of systems will be necessary to create, including the nominal behavior of the system and different faulty behaviors. In order to achieve high quality fault isolation, it is necessary to have complete and consistent models. To be able to use all applications that Rodon feature for a modeled system, preferable characteristics are that the model should be static, have discrete control signals, and have well defined system behavioral modes.

During development of a system Rodon can be used to improve and easy the work for failure analysis, guidance of sensor placements, evaluation of tests, generation of decision structures, and fault isolation. Since design of tests during development is a desirable application that Rodon does not have, two different methods are presented that utilizes Rodon to generate all possible limit checking tests.

In conclusion, Rodon can be very useful in several different aspects if introduced, but benefits gained by using Rodon will have to be compared to the labor cost of creating good models.

This work is a research in the field of systems science, emphasising the importance of applying models and theories that have been developed in this area. This work studies the possibility of using James Miller's living systems theory (LST) as a diagnostic tool. The application area was project management processes used when developing computerised information systems. The focus on the analyses was on the critical subsystems that process information. Based on this study it was found that LST function as a diagnostic tool according to the following criteria: it was possible to identify the critical subsystems in the application, the critical subsystems covered relevant information flows in the application and LST could make a unique contribution in the analyses of the application.

Technical systems of today are often complex and integrated. To maintain operational functionality and security it is sometimes necessary to have a surveillance system which can detect a fault in an early stage. The device that detects and locates the faulty component is called a diagnosis system. There are several different approaches to fault diagnosis, this study focus on a part of technical diagnosis that uses a model of the system for assistance to make the diagnosis. In the creation of the model of the system computer tools can be used, in this thesis it is investigated if one such software tool is practical to use in the building of the model of the system.

One aspect of technical diagnosis based on models is the creation of a model of the process in question with support for faults, a fault model. In this thesis, Modelica is used to create fault models. The models produced from Modelica are then analyzed by some existing diagnosis algorithms.

The approach of designing fault models in Modelica and then exporting the corresponding system and perform diagnosis analysis on it in Matlab is considered feasible. But the process falls a bit short of the aim of a automatic model building and and diagnosis analysis procedure.

If the results from this thesis are to be used in the future will depend on if the freedom that Modelica gives in designing fault models are deemed worthwhile to accommodate so that the results from the modeling suits the diagnosis analysis. One way to do ease this transformation of data is to restrict the freedom of model designs in Modelica but then some of the benefit that Modelica brings is negated. It is shown here however that it it possible to design fault models in Modelica and then use the results to analyze the models regarding sensor placement and test design.

AUTOSAR was created as a standard software infrastructure to be able to fulfill a very large amount of requirements. These days, more and more OEMs are trying to introduce AUTOSAR in their products. Since there are a large amount of diagnostic IDs needed in the Engine control unit and also a huge effort is necessary to configure the ECU, it is very much important to have a tool to generate some parts of the Engine Control Unit software, most importantly the diagnostics software. Diagnostic Communication manager is one such AUTOSAR module which deals with a huge amount of diagnostic data identifiers. Also at BEG, In the actual Non-AUTOSAR Bosch Automotive software, there are a number of different features that are needed and expected in the future AUTOSAR software. The aim of this thesis is to develop a tool that successfully introduces AUTOSAR in the BEG projects with all the necessary features and that is best in terms of Usability, Maintainability, and Improvability. This tool has to generate the complete AUTOSAR Diagnostic communication manager software with all the necessary features. The work can be divided into two parts. The first part includes a complete analysis of the existing tools that are used to generate configuration files and code. Then, List out all the possibilities of each tool, find their advantages and disadvantages and compare each of the tools, either individually or as a combination of tools. This is followed by documenting the choice of best way to generate AUTOSAR DCM with all the necessary features. In the second part, the implementation is carried out. After the best tool is chosen, the implementation of the features for that particular tool is planned accordingly so that it generates the DCM software. Implementation is made and then it is tested with the existing test bench.

The purpose of this research was to provide further evidence for a social expertise view of adult age differences in social cognitive functioning. Of specific interest was the extent to which such a perspective can be used to explain the differential use of trait-diagnostic information by young and older adults in the construction of social judgments. The use of such information has been shown to increase with age (Hess & Auman, 2001), suggesting the presence of superior social expertise in older adults when compared to younger adults. In the current research, factors associated with accessibility to relevant knowledge?extremity of trait-relevant behaviors and the amount of diagnostic information?was manipulated to determine if the differences between presumed experts (i.e., older adults) and nonexperts (i.e., younger adults) were attenuated when the salience of trait-diagnostic information was increased. Young, middle-aged, and older adults studied a series of behavioral description describing fictitious target individuals. Study times for individual behaviors contained in these descriptions and impression ratings for each target person were examined. Results of this study replicate past research; specifically, diagnostic information was studied longer and had a stronger impact on impression ratings than did nondiagnostic information, and the impact of diagnosticity increased with age. Further, extreme cues served to enhance the already present diagnostic effects in study time, while also causing ratings of target individuals to be more negatively rated overall. The expected moderation of age differences in the use of diagnostic information based on the extremity of cues did not follow the expected direction. The relationship between age, extremity of cues, and the use of diagnostic information was not significant, suggesting that extremity did not serve to differentially enhance the accessibility of knowledge structures across age groups as originally expected. In addition, larger amounts of diagnostic information actually resulted in greater age differences in the impact of diagnostic information on impression ratings. This, along with the absence of age differences when minimal diagnostic information was available, may suggest that those with expert knowledge are only willing to use it when sufficient cues are presented. I additionally tested an alternative explanation for observed age differences in the use of diagnostic information. Specifically, I investigated whether age differences in implicit beliefs regarding the stability of traits might mediate age differences. No support was obtained for this hypothesis. In sum, although the results were not entirely consistent with expectations, they were generally supportive of an aging-related increase in social expertise as an explanation for age differences in social judgments.

Over the past twenty years industrial communications networks have become common place in most industrial plants. The high availability of these networks is crucial in smooth plant operations. Therefore local and remote diagnostics of these networks is of primary importance in solving any existing or emerging network problems. Users for most part consider the “plant networks” as black boxes, and often not sure of the actual health of the networks. The major part of the work outlined in this research concentrates on the proposed “Network Diagnostics Model” for local and remote monitoring. The main objective of the research is to aid the establishment of tools and techniques for diagnosis of the industrial networks, with particular emphasis on PROFIBUS and PROFINET. Additionally this research has resulted in development of a number of devices to aid in network diagnostics. The work outlined in this submission contributes to the developments in the area of online diagnostics systems. The development work was conducted in the following phases: 1. Development of Function Block (FB) for diagnosing PROFIBUS network for implementation on PLC. 2. Development of OPC server for diagnosing PROFIBUS network for implementation on PC. 3. Development of a web based diagnostic software for multiple fieldbuses for implementation on imbedded XP platform. 4. Development of OPC server for diagnosing PROFINET network for implementation on PC 5. Conformance testing of masters (PLC) in PROFIBUS network to increase the health of the network. 6. Use of diagnostics tools for performance analysis of fieldbuses networks for high performance applications. The research work outlined in this submission has made a significant and coherent contribution to online diagnostics of fieldbus communications networks, and has paved the way for the introduction of the online diagnostics devices to the market place. It has shown that the proposed model provides a uniform framework for research and development of diagnostics tools and techniques for fieldbus networks. Organizations that use fieldbus should consider installing advanced online diagnostic systems to boost maintenance efficiency and reduce operating costs, and maintain the availability of plant resources. Based on the experience gained over a number of years a multilayer model is proposed for future development of diagnostics tools.

Information and communication technologies have been playing a crucial role in improving the efficiency and effectiveness of learning and teaching in higher education. Two decades ago, research studies were focused on how to use artificial intelligence techniques to imitate teachers or tutors in delivering learning sessions. Machine learning techniques have been applied in several research studies to construct a student model in the context of intelligent tutoring systems. However, the usage of intelligent tutoring systems has been very limited in higher education as most educational institutions are in favour of using virtual learning environments (VLEs). VLEs are computer-based systems that support all aspects of teaching and learning from provision of course materials to managing coursework. In this research study, the emphasis is on the assessment aspect of VLEs. A literature review revealed that existing computer-based formative assessments have never utilised unsupervised machine learning to improve their feedback mechanisms. Machine learning techniques have been applied to construct student models, which is represented as categories of knowledge levels such as beginning, intermediate and advanced. The student model does not specify what concepts are understood, the gap of understanding and misconceptions. Previously, a snap-drift modal learning neural network has been applied to improve the feedback mechanisms of computer-based formative assessments. This study investigated the application of snap-drift modal learning neural network for analysing student responses to a set of multiple choice questions to identify student groups. This research study builds on this previous study and its aim is to improve the effectiveness of the application of snap-drift modal learning neural network in modelling student responses to a set of multiple choice questions and to extend its application in modelling student responses gathered from object-oriented programming exercises. A novel method was proposed and evaluated using trials that improves the effectiveness of snap-drift modal learning neural network in identifying useful student group profiles, representing them to facilitate generation of diagnostic feedback and assigning an appropriate diagnostic feedback automatically based on a given student response. Based on the insight gained into the use of this novel method, we extend it to identify useful student group profiles that represent different programming abilities for writing an object-oriented class. The purpose of identifying student group profiles is to facilitate construction of diagnostic feedback that improves the development of basic object-oriented programming abilities. Overall, the main objectives of this research project were addressed successfully. New insights are gained into the application of unsupervised learning in general and snap-drift modal learning in particular. The proposed methods are capable of improving the feedback mechanisms of existing computer-based formative assessment tools. The improved computer-based formative assessments could have a huge impact on students in improving conceptual understanding of topics and development of basic object-oriented programming abilities.

L’évaluation des modèles hydrologiques est généralement basée sur des comparaisons des séries de débit observées et simulées à l’aide de critères de performance tels que le Nash-Sutcliffe Efficientcy. Bien que cette approach fournisse des mesures pertinentes de la performance prédictive d’un modèle, elle ne fournit que très peu d’informations sur les raisons d’une bonne ou d’une mauvaise performance. Gupta et al. (2008) ont proposé d’utiliser plutôt des signatures hydrologiques, des indicateurs qui caractérisent le comportement d’un bassin versant. Étant donné que les signatures hydrologiques peuvent être liées aux processus hydrologiques, la comparaison des signatures hydrologiques observées et simulées permet l’évaluation du modèle tout en offrant des diagnostics, i.e. des indications sur les processus hydrologiques qui y sont bien ou mal représentés.Dans cette thèse de doctorat, nous nous concentrons sur l’interprétation et le pouvoir diagnostique des signatures hydrologiques et comment celles-ci peuvent être utilisées pour guider l’amélioration d’un modèle distribué. Nous présentons la construction d’un jeu de signatures hydrologiques, utilisant uniquement des données largement disponibles – précipitations, débit et température de l’air – pour caractériser le fonctionnement hydrologique de 4 sous-bassins versants (Sud-Est de la France) et 10 bassins versants de montagne à influence nival (Southern Sierra, Californie, États-Unis). Des signatures hydrologiques existantes et des nouvelles sont sélectionnées et/ou développées. Collectivement, elles permettent de caractériser le comportement de bassins versants dans une grande variété de contextes hydro-climatiques. En nous concentrant sur les processus de neige, des mesures de neige supplémentaires nous permettent d’évaluer la pertinence des signatures hydrologiques dédiées aux processus nivaux. Par ailleurs, le modèle J2000 est déployé sur le bassin versant de l’Ardèche et une analyse de sensibilité est réalisée afin de comprendre comment les signatures hydrologiques sont liées aux paramètres du modèle. Cela nous permet de déterminer la façon dont elles doivent être interprétées dans le contexte du modèle J2000 de l’Ardèche et permet l’évaluation de leur pouvoir diagnostic. Enfin, en combinant les résultats de l’analyse de sensibilité avec des comparaisons entre signatures observées et simulées, nous entreprenons un diagnostic approfondi du modèle afin de dériver et tester des recommandations pour son amélioration. Nous identifions des déficiences du modèle, principalement liées au stockage de l’eau souterraine et des sols, mettant en évidence des problèmes de représentation spatiale des propriétés géologiques et pédologiques
The evaluation of hydrological models is typically based on comparisons of observed and sim-ulated streamflow time series using performance metrics such as the Nash-Sutcli˙e Eÿciency. Although it provides relevant measures of the predictive performance of a model, this type of approach provides very little information on the reasons behind good or bad performance. Instead, Gupta et al. (2008) proposed to use hydrological signatures which are indicators that characterize catchment behaviors. Because they can be related to hydrological processes, using them when comparing observation with simulation enable the evaluation of the model while o˙ering diagnostics, i.e. indications on the hydrological processes that are well or badly repre-sented in the model.In this PhD thesis, we focus on the interpretations and diagnostic power of hydrological signatures and how they can be used to guide the improvement of a distributed model. We present the building of a set of hydrological signatures, using only widely available data – pre-cipitation, streamflow and air temperature – to characterize the hydrological functioning of 4 Ardèche sub-catchments (South East of France) and 10 snow dominated catchments of the Southern Sierra mountains (California, USA). Already existing and new hydrological signatures are selected and/or designed. Collectively, they can characterize catchment behavior in a wide variety of hydro-climatic contexts. We demonstrate the value of additional snow measurements to evaluate the information content of snow dedicated hydrological signatures. In the context of the Ardèche catchment, we set up the J2000 distributed model and use a sensitivity analysis to understand how the hydrological signatures are linked to the model parameters. This provides insights into how they are to be interpreted in the context of the J2000 Ardèche model and allows the assessment of their diagnostic power. Finally, combining the results of the sensi-tivity analysis with comparisons between observed and simulated hydrological signatures, we undertake an in-depth diagnostic of the model to provide and test recommendations for its improvement. Deficiencies of the model functioning are identified, mainly related to soil and groundwater storage and fluxes, highlighting issues in the spatial representation of soil and geological properties

En réanimation, la démarche d’information de l’entourage d’un patient est souvent rendue difficile en raison de l’urgence, de l’absence des proches ou de proches injoignables, du principe de respect du secret médical et de la volonté du patient, et souvent de l’absence de l’interlocuteur légal représenté par la personne de confiance. Le cadre d’évolution légal actuel se révèle à l’usage en réanimation trop étroit pour satisfaire au besoin et à la nécessité de savoir des proches privés du contact avec leur parent hospitalisé et à la manifestation d’une empathie envers eux. A partir de la problématique de la démarche diagnostique en réanimation, du vivant du patient ou après son décès, nous proposons une évolution de la réflexion relative à la dispensation de l’information relative au patient auprès des proches permettant au médecin réanimateur d’évoluer plus aisément face au concept d’autonomie en privilégiant de façon temporaire le principe de bienfaisance tourné, de fait, vers le patient mais également vers ses proches
Information of relatives in intensive care unit is often difficult because of emergency, absence of relatives, respect of confidentiality and patient’s wishes and more often, because of absence of legal representative. French law does not provide for information of the next of kin in intensive care unit. We carried out two studies among intensivists on information about diagnostic processes during patient stay and immediatly after death. We suggested that a new approach of information could join respect of patient’s autonomy and beneficience towards relatives

This is a Master Thesis at the Computer Science Program at Växjö University. It has been made at BSR in Växjö where the development of a diagnostic tool has been in the running since the beginning of 2008. Previously there was a base developed within the company which I was to use for developing the two layers that I was responsible for – ProtocolLayer and PresentationLayer. In the end it should lead to a XML-based diagnostic tool. The technologies used have been .NET with its language C# and XML. The future purpose for this tool is for it to be used within the company to receive important information about the cars when developing other BSR products.

This thesis has focused on developing the tool to be working with the car manufacturer VOLVO and its protocols. The idea is that the diagnostic tool in the future could be extended to be working with all possible car manufacturers’ protocols and to enlighten the extensibility the project has been done using the product-line approach.

Prior research has indicated that laypeople construct mental representations of physical symptoms in order to attempt to understand illness (e.g., Leventhal, Safer, Panagis, 1983; Leventhal & Contrada, 1987; Lau, Bernard, & Hartman, 1989). These "illness representations" are influenced by prior experience with and prior knowledge about illness as well as efforts to seek additional information through social channels or media. More and more, the internet is a prominent source of health information, especially for older adults (aged 50 year and up). Yet, few studies have systematically examined how older adults search for health information online. Similarly, recent trends in healthcare such as health consumerism assume that patients will be more empowered if they have access to more information. However, little has been done to investigate whether patients, in fact, feel more empowered after acquiring online health information. The current study examined the online health information seeking of older adults (N = 79) in order to determine the cognitive and diagnostic processes that older adults use to acquire information. Older adults read a vignette which depicted one of two common illnesses and then were asked to "think-aloud" while they attempted to diagnose the illness. Older adults then diagnosed the illness using either a traditional search engine (Google) or popular self-diagnosis tool (WebMD Symptom Checker), and answered questions about illness representations, cognitive effort, web interactivity, and feelings of empowerment after the search. Quantitative results showed inconsistent change in illness representations. Plausible reasons for a lack of findings are discussed. Participants who used WebMD perceived greater cognitive effort while using the computer program than those who used Google, and participants who were inaccurate in their diagnosis perceived greater cognitive effort of diagnosing than those who were accurate. Accuracy was unrelated to perceived interactivity, age, or search method. Participants 50-64 years old found a new version of WebMD to be less interactive than Google. In contrast, participants 65 years or older perceived no difference in interactivity depending upon search method. In terms of empowerment, participants who used Google perceived greater choice than WebMD. There were no differences in feelings of competence depending upon search method. Qualitative results showed that participants spent the majority of time navigating the computer and processing health information. Most participants diagnosed the illness by eliminating diseases whose symptoms did not match the symptoms of the illness vignette. Participants tended to visit commercial health websites such as Everyday Health and begin their information search by typing a vignette symptom into the search bar. Participants who were 65 years or older were less confident about their diagnosis than 50-64 year old participants. Finally, participants who used Google to diagnose were more likely to comment about the credibility of the information found when compared to those who used WebMD. The current study found no change in illness representations after an online information search. However, this produces questions as to the amount of time in which the layperson constructs his/her illness representation. In addition, few differences in interactivity, accuracy, or empowerment were found between an online health information search conducted with a search engine as compared to a self-diagnosis tool. However, individual differences suggest that different age cohorts may prefer information to be presented in different ways which could influence web design. Further studies in human-computer interaction and health cognition may be able to answer the questions that arose.

In this dissertation we study the feature selection and classification problems and apply our methods to real-world medical and biological data sets for disease diagnosis. Classification is an important problem in disease diagnosis to distinguish patients from normal population. DAMIP (discriminant analysis -- mixed integer program) was shown to be a good classification model, which can directly handle multigroup problems, enforce misclassification limits, and provide reserved judgement region. However, DAMIP is NP-hard and presents computational challenges. Feature selection is important in classification to improve the prediction performance, prevent over-fitting, or facilitate data understanding. However, this combinatorial problem becomes intractable when the number of features is large. In this dissertation, we propose a modified particle swarm optimization (PSO), a heuristic method, to solve the feature selection problem, and we study its parameter selection in our applications. We derive theories and exact algorithms to solve the two-group DAMIP in polynomial time. We also propose a heuristic algorithm to solve the multigroup DAMIP. Computational studies on simulated data and data from UCI machine learning repository show that the proposed algorithm performs very well. The polynomial solution time of the heuristic method allows us to solve DAMIP repeatedly within the feature selection procedure. We apply the PSO/DAMIP classification framework to several real-life medical and biological prediction problems. (1) Alzheimer's disease: We use data from several neuropsychological tests to discriminate subjects of Alzheimer's disease, subjects of mild cognitive impairment, and control groups. (2) Cardiovascular disease: We use traditional risk factors and novel oxidative stress biomarkers to predict subjects who are at high or low risk of cardiovascular disease, in which the risk is measured by the thickness of the carotid intima-media or/and the flow-mediated vasodilation. (3) Sulfur amino acid (SAA) intake: We use 1H NMR spectral data of human plasma to classify plasma samples obtained with low SAA intake or high SAA intake. This shows that our method helps for metabolomics study. (4) CpG islands for lung cancer: We identify a large number of sequence patterns (in the order of millions), search candidate patterns from DNA sequences in CpG islands, and look for patterns which can discriminate methylation-prone and methylation-resistant (or in addition, methylation-sporadic) sequences, which relate to early lung cancer prediction.

In order to cope with system complexity and dynamic environments, modern industries are investing in a variety of sensor networks and data acquisition systems to increase information visibility. Multi-sensor systems bring the proliferation of high-dimensional functional Big Data that capture rich information on the evolving dynamics of natural and engineered processes. With spatially and temporally dense data readily available, there is an urgent need to develop advanced methodologies and associated tools that will enable and assist (i) the handling of the big data communicated by the contemporary complex systems, (ii) the extraction and identification of pertinent knowledge about the environmental and operational dynamics driving these systems, and (iii) the exploitation of the acquired knowledge for more enhanced design, analysis, monitoring, diagnostics and control. My methodological and theoretical research as well as a considerable portion of my applied and collaborative work in this dissertation aims at addressing high-dimensional functional big data communicated by the systems. An innovative contribution of my work is the establishment of a series of systematic methodologies to investigate the complex system informatics including multi-dimensional modeling, feature extraction and selection, model-based monitoring and root cause diagnostics. This study presents systematic methodologies to investigate spatiotemporal informatics of complex systems from multi-dimensional modeling and feature extraction to model-driven monitoring, fault identification and root cause diagnostics. In particular, we developed a multiscale adaptive basis function model to represent and characterize the high-dimensional nonlinear functional profiles, thereby reducing the large amount of data to a parsimonious set of variables (i.e., model parameters) while preserving the information. Furthermore, the complex interdependence structure among variables is identified by a novel self-organizing network algorithm, in which the homogeneous variables are clustered into sub-network communities. Then we minimize the redundancy of variables in each cluster and integrate the new set of clustered variables with predictive models to identify a sparse set of sensitive variables for process monitoring and fault diagnostics. We evaluated and validated our methodologies using real-world case studies that extract parameters from representation models of vectorcardiogram (VCG) signals for the diagnosis of myocardial infarctions. The proposed systematic methodologies are generally applicable for modeling, monitoring and diagnosis in many disciplines that involve a large number of highly-redundant variables extracted from the big data. The self-organizing approach was also innovatively developed to derive the steady geometric structure of a network from the recurrence-based adjacency matrix. As such, novel network-theoretic measures can be achieved based on actual node-to-node distances in the self-organized network topology.

Dissertação (mestrado) - Universidade Federal de Santa Catarina, Centro Tecnologico
Made available in DSpace on 2012-10-17T02:16:37Z (GMT). No. of bitstreams: 0Bitstream added on 2016-01-08T22:01:06Z : No. of bitstreams: 1 108977.pdf: 8642636 bytes, checksum: 22d84c7676dc99b8fd3da9175d46c411 (MD5)
Uma das principais dificuldades que apresentam os sistemas especialistas para apoio à tomada de decisão na área médica é que a imprecisão contida nos dados a serem utilizados é esquecida nos sistemas baseados em computador, fazendo-os diferentes da realidade. Neste trabalho, depois de uma revisão dos principais métodos para tratar com incerteza e imprecisão, é proposto um método para tratar a imprecisão nos dados clínicos utilizando a lógica nebulosa. O raciocínio baseado em casos é um paradigma simbólico da inteligência artificial para modelar o raciocínio do especialista. Este paradigma foi adotado para evitar a perda de informação intrínseca nos sistemas nebulosos baseados em regras. A teoria pesquisada é ilustrada pela implementação de um protótipo experimental no domínio da reumatologia.

The development of decision support systems for equipment diagnosis has been found to be an iterative process whereby functionality and knowledge are continually added to a prototype until satisfactory performance is achieved. In order to reduce both the dependency on compiled knowledge sources and the number of prototype stages necessary to develop diagnostic decision support systems, this thesis examines, adapts and applies a set theoretical approach to mechanism diagnosis first developed in the field of Artificial Intelligence. The approach does not require the development of computational models to simulate equipment behaviour.
The set theoretical approach was applied to the development of a diagnostic decision support system for a semi-automated Atlas Copco Wagner ST-8B Load-Haul-Dump vehicle. Hypothesis sets were generated for the vehicle's hydraulic circuit and Deutz FL-413-FW diesel engine. A high level of diagnostic resolution was achieved for the hydraulic circuit, but limited resolution was achieved for the diesel engine. This was postulated to be due to the ratio of observable system outputs to input sub-systems, and the number of least repairable units making up each system.
Manual knowledge acquisition was undertaken in an underground mine to refine the diagnostic knowledge developed from the hypothesis sets and to add knowledge to discriminate between competing failure hypotheses. Heuristic failure likelihoods were used to rank hypotheses in order of frequency of occurrence. The knowledge base was implemented as a hypertext decision support system using HyperText Mark-up Language (HTML). The resulting decision support system is platform independent, upgradeable and able to be maintained by site personnel. The system is currently installed at surface level and at 1800 level at INCO Limited's Stobie Mine in Sudbury, Ontario.
The thesis makes a number of original contributions, the first two of which are of generic significance. It is the first work to apply set theoretical concepts to structural models of mobile mining equipment in order to diagnose faults. A number of modifications are advanced to the conventional trace-back analysis technique for generating contributor and normality sets, and heuristic guidelines are provided for estimating the costs and benefits of developing, implementing and maintaining diagnostic decision support systems. It is also the first work to formalise a decision support system in HTML and to suggest the application of company-wide internets ("intranets") to disseminate maintenance knowledge within mines.

Dissertação (mestrado) - Universidade Federal de Santa Catarina, Centro de Ciencias da Saude
Made available in DSpace on 2012-10-16T06:51:37Z (GMT). No. of bitstreams: 0
O sistema baseado em conhecimentos foi desenvolvimento em dois módulos. O módulo de triagem e o de classificação. O módulo de triagem é construído adaptando o questionário de Fernandes (1993), elaborado e validado em Porto Alegre, através de um inquérito populacional. Este módulo de classificação utiliza as técnicas de inteligência artificial simbólica, com a programação orientada a objetos, usando shell Kappa, classificando as crises epilépticas baseada nos critérios da ILAE "International League Against Epilepsy" (1981). A estrutura do sistema assemelha-se ao desenho dos estudos epidemiológicos populacionais realizados em dois estágios. O sistema foi testado e demonstrou-se um bom desempenho. Os aspectos atuais de avaliação e as implicações ético-legais também foram discutidos. O sistema implementado pode ser útil na pesquisa clínico epidemiológica, no ensino da epilepsia e também pode servir de base para o desenvolvimento de outros sistemas.

Dissertação (mestrado) - Universidade Federal de Santa Catarina, Centro Tecnológico. Programa de Pós-Graduação em Ciência da Computação.
Made available in DSpace on 2013-07-16T01:36:27Z (GMT). No. of bitstreams: 1 222903.pdf: 10700721 bytes, checksum: febbe9a0a9cd44dffbe0ca92a0187d2e (MD5)
Devido à grande variabilidade anatômica relativa ao cérebro do ser humano, o seu estudo implica uma área incerta da ciência, e esforços direcionados ao objetivo de gerar um mapeamento dessa complexa estrutura obtêm constantes sucessos, mas também fracassos. Atualmente, com a crescente capacidade computacional e com a disponibilidade de metodologias para alinhamento de imagens, novos passos direcionados ao esforço de mapear o cérebro humano estão sendo dados. A existência de um "cérebro médio", que seja estatisticamente representativo em relação às características de uma dada população, pode vir a auxiliar o acompanhamento da evolução de doenças degenerativas cerebrais, ao fornecer uma base sólida para identificação de anomalias, assim como para servir de base para a tomada de decisão clínica. Este trabalho apresenta o resultado de dois anos de pesquisa e desenvolvimento junto ao projeto Cyclops, no intuito da elaboração de uma metodologia de registro volumétrico de imagens médicas e conseqüente geração de mapas médios do cérebro humano. Como fator motivador deste trabalho desenvolveu-se uma metodologia para a avaliação do grau de atrofia global e local do córtex cerebral, com o intuito de auxiliar os médicos neurorradiologistas no processo de diagnóstico precoce da doença de Alzheimer, assim como no acompanhamento e predição da evolução estatisticamente representativa do estado de MCI- Mild Cognitive Inparment para a doença de Alzheimer. Ao final da pesquisa obteve-se: a) descrição metodológica de geração de cérebros médios. b) uma ferramenta para registro de imagens mono-modais/multi-subject; c) ferramenta para geração de cérebros médios; d) ferramenta de geração de mapas de cores para salientar regiões atrofiadas do córtex; e) todas as funcionalidades discutidas neste trabalho sob a forma de um framework extensível.

As machines are increasingly used to fulfill even more needs of mankind, the dependence upon those machines increase. To prevent catastrophic failure and to facilitate maintenance a diagnostic system can be used. A diagnostic system supervises the system and can alarm the operator when a fault has occurred, and possibly determine what the cause may be. One architecture of a diagnostic system is a number of tests run by an on-board computer checking certain combinations of sensor values and control signals chosen in advance. To design these tests is a difficult task, which leads to the desire to automate the test construction. A part of this task can be performed using structural methods.

In this thesis model based diagnosis is considered. This means that a formal mathematical model is used. The models typically consist of a number of equations describing the behavior of the system. In structural methods it is only considered if a variable exists in an equation or not. The goal of this master thesis project has been to apply structural methods to RODON models. RODON is a software diagnostics tool brought to market by Sörman Information & Media, which can perform various diagnostic-related tasks based on a single model. This model is defined in an object oriented fashion using a Modelica-like language called Rodelica. A prototype implementation of a structural algorithm plug-in has been developed and integrated into RODON. An additional part of the project has been to investigate further possible uses of structural algorithms in RODON, apart from diagnostic test construction. This has been performed as a series of interviews with Sörman and university employees.

The work performed in this thesis has shown that it is possible to apply structural methods to RODON models. It has also shown that even a prototype implementation can handle quite large systems. Some problems have been found as well, most notably in extracting a structural model from a RODON model. A consequence is that the developed structural plug-in only works for a subset of RODON models. It might be possible to deal with these problems if more time would be spent on the task. Finally, the interview survey revealed other possible uses of structural methods in RODON, including optimal sensor placement analysis and isolability and detectability analysis.

La contribución que se pretende con esta tesis se refiere a la implantación de un sistema de diagnosis de fallos en plantas químicas completas integrado al sistema de supervisión, gestión y control de la producción.
El sistema de diagnosis de fallos que se presenta consiste en una combinación de un sistema de reconocimiento de patrones basado en redes neuronales artificiales y un sistema de inferencia basado en la lógica difusa. La información necesaria para desarrollar el sistema de diagnosis incluye los datos históricos, un análisis de riesgo y operabilidad y un modelo de la planta química. La entrada al sistema son las mediciones directas o indirectas de la planta y la salida consiste en una señal para cada fallo (0: no fallo; 1: fallo). Primero se definen los fallos posibles. La red neuronal se entrena con datos históricos de fallos ocurridos en el pasado, con el objeto de reconocer los patrones respectivos. En el caso de que no se posean los datos históricos de alguno de los fallos, por ejemplo porque nunca hayan ocurrido, se obtienen los patrones mediante la simulación, usando el modelo de la planta. El sistema de lógica difusa contiene un conjunto de reglas si-entonces que pueden ser de dos tipos: las basadas en el conocimiento de la planta, mediante el análisis de riesgo o la experiencia con la simulación, y las basadas en la experiencia con el uso de la red neuronal, previamente entrenada. Otro aspecto novedoso es la posibilidad de entrenar la red neuronal con "características" extraídas de las variables medidas mediante su pre-procesamiento con wavelets. Esta variante permite obtener un alto rendimiento del sistema de diagnosis en plantas químicas discontinuas y continuas complejas. Para optimizar los parámetros de los componentes del sistema de diagnosis se propone un índice de rendimiento. Además, se utiliza el índice de rendimiento para comparar el sistema propuesto con otros métodos.
Las señales dadas por el sistema de diagnosis pueden ser usadas por el sistema de programación de la producción para actualizar el plan de la manera más efectiva, por el sistema de control para actuar en forma automática y por los operadores de planta como soporte para la toma de decisiones. Se han sentado las bases para la traducción de la salida del sistema de diagnosis para su utilización por los demás niveles del soporte informático. Se usa una estrategia basada en el análisis de riesgo y operabilidad de la planta.
El sistema propuesto es consecuencia de sucesivas mejoras, al trabajar con diferentes casos de estudio. Los escenarios académicos corresponden a una planta química continua con una corriente de reciclo y un reactor discontinuo. Los casos a escala de planta piloto corresponden a escenarios construidos en la UPC: un reactor de gasificación, un reactor semicontinuo y una planta discontinua multipropósito. Los escenarios industriales corresponden a dos refinerías de azúcar y a un sector de una planta petroquímica.
El sistema propuesto muestra ser ventajoso respecto a otros métodos tanto en cuanto a la rapidez de diagnosis como en cuanto a su capacidad para aislar los fallos. La simplicidad del desarrollo y la flexible estrategia de implementación del sistema propuesto auguran un futuro promisorio a la tecnología presentada. Nuevas líneas de investigación se pueden emprender mediante el desarrollo de un sistema de gestión de las alarmas. Otro aspecto importante es la posibilidad de participar en la estandarización de las interfaces del programa de diagnosis.
The pretended contribution of this thesis deals with the implementation of a fault diagnosis system in chemical plants integrated to the monitoring, management and control system. The proposed fault diagnosis system consists in a combination of a pattern recognition approach based on artificial neural networks and an inference system based on fuzzy logic. The information needed to develop the fault diagnosis system includes the historical data, the hazard and operability study and the model of the chemical plant. The inputs to the system are the direct or indirect measurements from the plant and the output consists in a signal for each fault (0: no fault; 1: fault). First, the possible faults are defined. The artificial neural network is trained with historical data of faults occurred in the past, with the aim of recognising the respective patterns. In the case that the corresponding historical data are not available, for example due to the no occurrence of the fault, the patterns are obtained through simulation, using the plant model. The fuzzy logic system contains a set of if-then rules that can be of two types: those based in the process knowledge, by the hazard analysis or by the experience with simulation, and those based on the experience with the use of an artificial neural network, previously trained. Other novel aspect is the possibility of artificial neural network training by using signals features that are extracted by its pre-processing using wavelets. This alternative allows a higher fault diagnosis system performance in batch and complex continuous chemical plants. In order to optimise the parameters of the components of the fault diagnosis system, a performance index is proposed. The performance index is also utilised to compare the proposed fault diagnosis system against other methods.
The signals provided by the fault diagnosis system can be used by the scheduling system to update the schedule in the most effective way, by the control system to take automated control actions and by plant's operators as support for decision-making. The basis of the translation of the system output, for its utilisation at other levels in the information system, has been settled.
The proposed strategy is based on the hazard and operability analysis.
The proposed system is the result of successive improvements, by working with different case studies. The academic scenarios correspond to a continuous chemical plant with a recycle stream and a batch reactor. The pilot plant scale cases correspond to scenarios built at UPC: a reactor gasifier, a fed-batch reactor and a multipurpose batch chemical plant. The industrial scenarios correspond to two sugar refineries and a sector of a petrochemical plant.
The proposed system shows to be advantageous with respect to other methods in relation to the fastness of the diagnosis and also its capacity to isolate faults. The simplicity of the development and the flexible strategy of implementation of the proposed fault diagnosis system give a promising future to the presented technology. New research lines can be considered by developing the alarm handling system. Other important aspect is the possibility of the participation in the standardisation of the interfaces of the fault diagnosis program.

As space exploration vehicles travel deeper into space, their distance to earth increases.The increased communication delays and ground personnel costs motivatea migration of the vehicle health management into space. A way to achieve thisis to use a diagnosis system. A diagnosis system uses sensor readings to automaticallydetect faults and possibly locate the cause of it. The diagnosis system usedin this thesis is a model-based reasoning tool called RODON developed by UptimeSolutions AB. RODON uses information of both nominal and faulty behavior ofthe target system mathematically formulated in a model.The advanced diagnostics and prognostics testbed (ADAPT) developed at theNASA Ames Research Center provides a stepping stone between pure researchand deployment of diagnosis and prognosis systems in aerospace systems. Thehardware of the testbed is an electrical power system (EPS) that represents theEPS of a space exploration vehicle. ADAPT consists of a controlled and monitoredenvironment where faults can be injected into a system in a controlled manner andthe performance of the diagnosis system carefully monitored. The main goal of thethesis project was to build a model of the ADAPT EPS that was used to diagnosethe testbed and to generate decision trees (or trouble-shooting trees).The results from the diagnostic analysis were good and all injected faults thataffected the actual function of the EPS were detected. All sensor faults weredetected except faults in temperature sensors. A less detailed model would haveisolated the correct faulty component(s) in the experiments. However, the goal wasto create a detailed model that can detect more than the faults currently injectedinto ADAPT. The created model is stationary but a dynamic model would havebeen able to detect faults in temperature sensors.Based on the presented results, RODON is very well suited for stationary analysisof large systems with a mixture of continuous and discrete signals. It is possibleto get very good results using RODON but in turn it requires an equally goodmodel. A full analysis of the dynamic capabilities of RODON was never conductedin the thesis which is why no conclusions can be drawn for that case.

 

La réduction du risque volcanique dans les départements ultramarins français s'inscrit dans une réflexion plus globale de gestion visant la caractérisation des enjeux et la diminution de leur vulnérabilité. Les îles de la Guadeloupe, la Martinique et la Réunion sont exposées à une multitude de risques naturels de par leur situation géographique et leur caractère insulaire. S'appuyant sur une approche comparative, aussi bien spatiale que temporelle, cette thèse de géographie propose une analyse des rapports des hommes à leur territoire marqué par la présence d'une menace latente : le volcanisme. Dans cette perspective, le risque est appréhendé, non comme un élément extérieur, mais comme un processus de construction sociale. A partir des scénarios éruptifs envisagés par les experts, ce diagnostic évalue, en amont de toute crise, les connaissances, les représentations et les comportements annoncés des populations exposées au risque. Il envisage les déterminants influents, tels le type de volcanisme, l'histoire éruptive ou l'identité des sociétéscréoles. Recenser les distorsions et cerner les réponses inadaptées permet d'envisager des outils d'information préventive plus efficients. Dans la lignée de travaux s'attachant à des thématiques similaires, cette recherche met en exergue la dimension de la composante humaine au coeur de la gestion des risques qu'on a qualifié un peu trop rapidement de "naturels". Au final, nous pointons les atouts à renforcer, les méconnaissances ou croyances à rectifier et les caractéristiques immuables des sociétés créoles, à prendre en compte pour une gestion de crise volcanique optimisée dans le futur.

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, Technology and Policy Program, 2016.
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 77-79).
During the mid-1960s, India came to the brink of an acute food crisis in the midst of heavy dependence on food imports. A period of rapid agricultural modernization that followed, known as Green Revolution, transformed India from a net importer of food into an exporter. Although an appropriate response for abating the impending starvation, the Green Revolution inflicted several unintended consequences. For example, regulatory structure and fertilizer subsidies for urea that were designed to stimulate growth instead resulted in a lock-in, which in turn incentivized vast over-fertilization across the country. Today, this is a well-recognized problem, and the Government of India has announced policies and schemes such as the National Soil Health Card Scheme to increase knowledge of soil condition and curb fertilizer use. In reality, however, the current need for information on soil health far exceeds the capacity for soil testing, highlighting the need for a radical approach to meeting this policy objective. This project, undertaken in collaboration with MIT Mechanical Engineering, takes a two-part approach to addressing this problem, with the design of a point-of-use soil testing sensor and an accompanying recommendation generation engine. This thesis presents the design of the latter based upon the answer to the following question: what constitutes an actionable information for resource constrained farmers? To answer it, we use a mixed methodology approach comprising (i) a combination of stakeholder interviews and design workshops to elicit user needs, and (ii) controlled experimentation with over 200 farmers covering an entire village to measure the actionability of information in soil health recommendations. The results of the analysis of experimental data reveal that the actionability of recommendations varies significantly within the population of farmers tested, and can be attributed to the level of information provided, the environment in which a farmer receives a recommendation, gender, and education level. Consequently, an effective point-of-use diagnostic system must adjust for these factors in order to maintain high actionability. To that end, we then use the experimental results to design a recommendation generation engine, the core of which is a soil health database that maximizes the actionability of information for a resource constrained farmer.
by Soumya Braganza.
S.M. in Technology and Policy

Ambulatory care comprises a major and increasingly important part of the U.S. and other countries’ health care sectors. Every year in the U.S., about 80% of the population seeks care at a doctor’s office, amounting to one billion visits. These visits divide almost equally between primary care and specialty clinic organizations. Diagnostic work is part of most ambulatory care, and central to over 40% of patient visits that originate due to a new problem or a flare-up of an ongoing chronic problem. Yet, the risks associated with diagnostic failures have not garnered much attention from health care leaders and policy makers until a recent National Academy of Medicine (NAM 2015) report synthesized research data with the statement that “most people will experience at least one diagnostic error in their lifetime, sometimes with devastating consequences.” This dissertation first reviews organizational theories and measurement challenges relevant to diagnostic safety and quality in the context of ambulatory care, and then presents three papers analyzing specific organizational factors hypothesized to enable or thwart an accurate and timely diagnosis. The first paper targets delayed diagnosis from missed evidence-based monitoring in high-risk conditions (e.g., cancer) within five specialty clinics in an urban publicly funded health system. The second paper analyzes staff-reported office problems that could lead to diagnostic error (e.g., not having test results when needed) in over 900 primary and specialty clinics across the nation. The third paper examines the associations between two types of time pressure (i.e., encounter-level and practice-level), organizational factors, and patient effects including perceptions of missed diagnostic opportunities. The three primary conclusions from this work are 1) organizational vulnerabilities for missed monitoring common to the different clinics included challenges with data systems, communications handoffs, population-level tracking, and patient activities, leading to the development of ‘design seeds’ for context-flexible solutions to improve diagnostic quality; 2) two organizational factors—stage of health information technology (HIT) deployment and patient safety culture are associated with diagnostic-related office problems, and 3) encounter and practice-level time stressors in primary care clinics are associated with perceptions of greater adverse effects on diagnosis and treatment, and worse patients’ experiences of chronic care from the clinic team, respectively, as well as associated with several organizational factors including HIT, patient-centered culture, relational coordination for interdependent teamwork, and leadership facilitation of changes to address frontline practice challenges. Taken together, the dissertation papers also demonstrate the applicability of the NAM Improving Diagnosis Conceptual Framework for research on ambulatory care organizations.

This dissertation is comprised of a collection of five distinct research projects which apply, evaluate and extend common methods for land surface data assimilation. The introduction of novel diagnostics and extensions of existing algorithms is motivated by an example, related to estimating agricultural productivity, of failed application of current methods. We subsequently develop methods, based on Shannon's theory of communication, to quantify the contributions from all possible factors to the residual uncertainty in state estimates after data assimilation, and to measure the amount of information contained in observations which is lost due to erroneous assumptions in the assimilation algorithm. Additionally, we discuss an appropriate interpretation of Shannon information which allows us to measure the amount of information contained in a model, and use this interpretation to measure the amount of information introduced during data assimilation-based system identification. Finally, we propose a generalization of the ensemble Kalman filter designed to alleviate one of the primary assumptions - that the observation function is linear.