Dissertations / Theses on the topic 'Personalized medicine support systems'

Advanced information technologies promise a massive influx of individual-specific medical data. These rich sources offer great potential for an increased understanding of disease mechanisms and for providing evidence-based and personalized clinical decision support. However, the size, complexity, and biases of the data pose new challenges, which make it difficult to transform the data to useful and actionable knowledge using conventional statistical analysis. The so-called “Big Data” era has created an emerging and urgent need for scalable, computer-based data mining methods that can turn data into useful, personalized decision support knowledge in a flexible, cost-effective, and productive way. The goal of my Ph.D. research is to address some key challenges in current clinical deci-sion support, including (1) the lack of a flexible, evidence-based, and personalized data mining tool, (2) the need for interactive interfaces and visualization to deliver the decision support knowledge in an accurate and effective way, (3) the ability to generate temporal rules based on patient-centric chronological events, and (4) the need for quantitative and progressive clinical predictions to investigate the causality of targeted clinical outcomes. The problem statement of this dissertation is that the size, complexity, and biases of the current clinical data make it very difficult for current informatics technologies to extract individual-specific knowledge for clinical decision support. This dissertation addresses these challenges with four overall specific aims: Evidence-Based and Personalized Decision Support: To develop clinical decision support systems that can generate evidence-based rules based on personalized clinical conditions. The systems should also show flexibility by using data from different clinical settings. Interactive Knowledge Delivery: To develop an interactive graphical user interface that expedites the delivery of discovered decision support knowledge and to propose a new visualiza-tion technique to improve the accuracy and efficiency of knowledge search. Temporal Knowledge Discovery: To improve conventional rule mining techniques for the discovery of relationships among temporal clinical events and to use case-based reasoning to evaluate the quality of discovered rules. Clinical Casual Analysis: To expand temporal rules with casual and time-after-cause analyses to provide progressive clinical prognostications without prediction time constraints. The research of this dissertation was conducted with frequent collaboration with Children’s Healthcare of Atlanta, Emory Hospital, and Georgia Institute of Technology. It resulted in the development and adoption of concrete application deliverables in different medical settings, including: the neuroARM system in pediatric neuropsychology, the PHARM system in predictive health, and the icuARM, icuARM-II, and icuARM-KM systems in intensive care. The case studies for the evaluation of these systems and the discovered knowledge demonstrate the scope of this research and its potential for future evidence-based and personalized clinical decision support.

Multiple Myeloma (MM) is a treatable, yet incurable, malignancy of bone marrowplasma cells. This cancer affects many patients and many succumb to relapse of tumor burden despite a large number of available chemotherapeutic agents developed for therapy. This is because MM tumors are heterogeneous and receive protection from therapeutic agents by the microenvironment and other mechanisms including homologous MM-MM aggregation. Therefore, therapy failure and frequent patient relapse is due to the evolution of drug resistance, not a lack of available drugs. To analyze and understand this problem, the evolution of drug resistance has been explored and presented herein. We seek to describe the methods through which MM cells become resistant to therapy, and how this resistance evolves throughout a patient’s treatment history. We achieve this in five steps. First we review the patient’s clinical history, including treatments and changes in tumor burden. Second, we trace the evolutionary tree of sub-clones within the tumor burden using standard of care fluorescence in situ hybridization (FISH). Thirdly, immunohistochemistry slides are stained and aligned to quantify the level of environmental protection received by surrounding cells and plasma in the bone marrow microenvironment (coined environment mediated drug resistance score [EMDR]). The fourth analysis type is produced through a novel 384-well plate ex vivo chemosensitivity assay to quantify sensitivity of primary MM cells to chemotherapeutic agents and extrapolate these findings to 90-day clinical response predictions. In addition to direct clinical application in the choice of best treatment, this tool was also used to study changes in sensitivity of patient tumors to other drugs, and it was observed that, upon relapse, in addition to developing resistance to the current line of therapy, tumors become cross-resistant to agents that they were never exposed to. Finally, MM-MM homologous aggregation is quantified to assess the level of drug resistance contributed by clustering of patient tumor cells, which causes upregulation of Bcl-2 expression and other resistance mechanisms1. The findings of such experimentation improve comprehension of the driving factors that contribute to drug resistance evolution on a personalized treatment basis. The aforementioned factors all contribute in varying degrees for unique patient cases, seven of which are presented in depth for this project. In summary: Environmental protection plays a critical initial role in drug resistance, which is followed by increase in tumor genetic heterogeneity as a result of mutations and drug-induced Darwinian selection. Eventually, environment-independent drug resistant subpopulations emerge, allowing the tumor to spread to unexplored areas of the bone marrow while maintaining inherited drug resistant phenotype2. It is our hope that these findings will help in shifting perspective regarding optimal management of MM by finding new therapeutic procedures that address all aspects of drug resistance to minimize chance of relapse and improve quality of life for patients.

La sécurité est une première étape essentielle dans l'amélioration de la qualité des soins. Devant l’importance de ces erreurs qui menacent 12 millions de patients aux USA chaque année ; plusieurs travaux ont essayé de trouver des solutions pour réduire les erreurs médicales et les effets indésirables des médicaments : La médecine basée sur la preuve, la médecine personnalisée et les systèmes d’aide à la décision médicale. Les sociétés savantes élaborent de façon périodique des recommandations de bonnes pratiques pour aboutir à instaurer une médecine basée sur la preuve. Ces recommandations sont considérées comme des outils efficaces pour faire pont entre la pratique médicale des praticiens et les preuves scientifiques.La simple diffusion de GBPC n’a qu’un impact limité sur les pratiques cliniques.Plusieurs études ont montré que l’informatisation de ces guides en les intégrant dans le Workflow clinique permet d’améliorer l’adhérence des médecins à ces recommandations. Les guides de bonnes pratiques cliniques ne couvrent pas les caractéristiques individuelles des patients. Un des objectifs d’amélioration des soins et de la réduction des effets indésirables des patients est la personnalisation de la prise en charge. Cette personnalisation nécessite l’utilisation de toutes les informations (cliniques, biologiques, génétiques, radiologiques, sociales..) pour caractériser le profil du patient. Nous avons développé une méthode de raisonnement hybride, CBFCM, capable d’utiliser des connaissances et des données hétérogènes. L’implémentation de la méthode a été faite avec des outils du web sémantique. Nous avons développé un environnement Open Source pour la modélisation et la formalisation des connaissances médicales (recommandations..). Nous avons validé la méthode avec plusieurs études dans le domaine des infections urinaires mais aussi dans d’autres domaines (pneumologie, stéatose hépatique non alcoolique, diabète gestationnel..). L’intégration des données génétiques, cliniques et biologiques nous a permis d’améliorer la prédiction de certaines maladies (NASH)
Several studies have tried to find ways to reduce medical and adverse drug errors:The evidence-based medicine, personalized medicine and clinical decision support systems. Many recommandations are developped periodically to improve a best practices. These recommendations are considered effective tools to bridge between medical practitioners and practice of scientific evidence. The use of the Clinical Practice Guidelines has a limited impact on clinical practice. Several studies showed that the computerization of these guides by integrating them into the clinical workflow improves adherence of physicians to these recommendations.One of the aims of improving care and reducing adverse effects of patients is personalizing care. This customization requires the use of all the information (clinical, biological, genetic, radiological, social..) to characterize the profile of the patient.We have developed a method of hybrid reasoning "Case Based Fuzzy CognitiveMaps" able to use knowledge and heterogeneous data. The implementation of themethod was made with semantic web technologies. We have developed an open source environment for modeling and formalization of medical knowledge.We validated the method with several studies in the field of urinary tract infections,but also in other areas (respiratory, nonalcoholic fatty liver disease, gestational diabetes..). The integration of genetic, clinical and laboratory data have allowed us to improve the prediction of certain diseases (NASH)

This work is motivated by the need of providing patients with a decision support system that facilitates the selection of the most appropriate treatment strategy in cancer treatment. Treatment options are currently subject to predetermined clinical pathways and medical expertise, but generally, do not consider the individual patient characteristics or preferences. Although genomic patient data are available, this information is rarely used in the clinical setting for real-life patient care. In the area of personalized medicine, the advancement in the fundamental understanding of cancer biology and clinical oncology can promote the prevention, detection, and treatment of cancer diseases. The objectives of this research are twofold. 1) To develop a patient-centered decision support model that can determine the most appropriate cancer treatment strategy based on subjective medical decision criteria, and patient's characteristics concerning the treatment options available and desired clinical outcomes; and 2) to develop a methodology to organize and analyze gene expression data and validate its accuracy as a predictive model for patient's response to radiation therapy (tumor radiosensitivity). The complexity and dimensionality of the data generated from gene expression microarrays requires advanced computational approaches. The microarray gene expression data processing and prediction model is built in four steps: response variable transformation to emphasize the lower and upper extremes (related to Radiosensitive and Radioresistant cell lines); dimensionality reduction to select candidate gene expression probesets; model development using a Random Forest algorithm; and validation of the model in two clinical cohorts for colorectal and esophagus cancer patients. Subjective human decision-making plays a significant role in defining the treatment strategy. Thus, the decision model developed in this research uses language and mechanisms suitable for human interpretation and understanding through fuzzy sets and degree of membership. This treatment selection strategy is modeled using a fuzzy logic framework to account for the subjectivity associated to the medical strategy and the patient's characteristics and preferences. The decision model considers criteria associated to survival rate, adverse events and efficacy (measured by radiosensitivity) for treatment recommendation. Finally, a sensitive analysis evaluates the impact of introducing radiosensitivity in the decision-making process. The intellectual merit of this research stems from the fact that it advances the science of decision-making by integrating concepts from the fields of artificial intelligence, medicine, biology and biostatistics to develop a decision aid approach that considers conflictive objectives and has a high practical value. The model focuses on criteria relevant to cancer treatment selection but it can be modified and extended to other scenarios beyond the healthcare environment.

Introduction: Intensive Care Unit (ICU) medical personnel, in an ongoing process termed ventilation management, utilize patient physiology and pathology data to define ventilator apparatus settings. Aims: The aim of the research is to develop and evaluate in comparison hybrid ventilation advisor systems, that could support ventilation management process, specific to lung pathology for patients ventilated in control mode. Methodology: A questionnaire was designed and circulated to Intensivists. Patient data, as defined by the questionnaire analysis, were collected and categorized into three lung pathologies. Three ICU doctors evaluated correlation analysis of the recorded data. Evaluation results were used for identifying models basic architecture. Two custom software toolboxes were developed for developing hybrid systems; namely the EVolution Of Fuzzy INference Engines (EVOFINE) and the FUzzy Neural (FUN) toolbox. Eight hybrid systems developed with EVOFINE, FUN, ANFIS and ANN techniques were evaluated against applied clinical decisions and patient scenarios. Results: Seventeen (17) models were designed for each of the eight (8) modeling techniques. The modelled process consisted of twelve physiology variables and six ventilator settings. The number of models’ inputs ranged from single to six based on correlation and evaluation findings. Evaluation against clinical recommendations has shown that ANNs performed better; mean average error as percentage for four of the applied techniques was 0.16%, 1.29% & 0.62 for ANN empirical, 0.05%, 2.23% & 2.30% for ANFIS, 0.93%, 2.33% & 1.89% for EVOFINE and 0.73%, 2.63% & 6.56 for FUN NM, in Normal, COPD and ALI-ARDS categories respectively. Additionally evaluation against clinical disagreement SD has shown that 70.6% of the NN empirical models were performing in 90% of their suggestions within clinical SD, while the percentages were 53%, 53% and 59% for the EVOFINE, ANFIS and NN Normalized models respectively. The EVOFINE and ANFIS produced Fuzzy Systems whose architecture is transparent for the user. Visual observation of ANFIS architectures revealed possibly hazardous advices. Evaluation against clinical disagreement has shown that the NN empirical was not producing hazardous advices, while EVOFINE, ANFIS and NN Normalized were shown to produce potentially hazardous advice in 17.6%, 23% and 5.8% of the developed models.

This thesis explored the role of health informatics (decision support systems) in caring for hospitalised patients with diabetes through a systematic review and by analysing data from University Hospital Birmingham, UK. Findings from the thesis: 1) highlight the potential role of computerised physician order entry system in improving guideline based anti-diabetic medication prescription in particular insulin prescription, and their effectiveness in contributing to better glycaemic control; 2) quantify the occurrence of missed discharge diagnostic codes for diabetes using electronic prescription data and suggests 60% of this could be potentially reduced using an algorithm that could be introduced as part of the information system; 3) found that hypoglycaemia and foot disease in hospitalised diabetes patients were independently associated with higher in-hospital mortality rates and longer length of stay; 4) quantify the hypoglycaemia rates in non-diabetic patients and proposes one method of establishing a surveillance system to identify non diabetic hypoglycaemic patients; and 5) introduces a prediction model that may be useful to identify patients with diabetes at risk of poor clinical outcomes during their hospital stay.

The idea of developing automated tools able to deal with the complexity of clinical information processing dates back to the late 60s: since then, there has been scope for improving medical care due to the rapid growth of medical knowledge, and the need to explore new ways of delivering this due to the shortage of physicians. Clinical decision support systems (CDSS) are able to aid in the acquisition of patient data and to suggest appropriate decisions on the basis of the data thus acquired. Many improvements are envisaged due to the adoption of such systems including: reduction of costs by faster diagnosis, reduction of unnecessary examinations, reduction of risk of adverse events and medication errors, increase in the available time for direct patient care, improved medications and examination prescriptions, improved patient satisfaction, and better compliance to gold-standard up-to-date clinical pathways and guidelines. Logistic regression is a widely used algorithm which frequently appears in medical literature for building clinical decision support systems: however, published studies frequently have not followed commonly recommended procedures for using logistic regression and substantial shortcomings in the reporting of logistic regression results have been noted. Published literature has often accepted conclusions from studies which have not addressed the appropriateness and accuracy of the statistical analyses and other methodological issues, leading to design flaws in those models and to possible inconsistencies in the novel clinical knowledge based on such results. The main objective of this interdisciplinary work is to design a sound framework for the development of clinical decision support systems. We propose a framework that supports the proper development of such systems, and in particular the underlying predictive models, identifying best practices for each stage of the model’s development. This framework is composed of a number of subsequent stages: 1) dataset preparation insures that appropriate variables are presented to the model in a consistent format, 2) the model construction stage builds the actual regression (or logistic regression) model determining its coefficients and selecting statistically significant variables; this phase is generally preceded by a pre-modelling stage during which model functional forms are hypothesized based on a priori knowledge 3) the further model validation stage investigates whether the model could suffer from overfitting, i.e., the model has a good accuracy on training data but significantly lower accuracy on unseen data, 4) the evaluation stage gives a measure of the predictive power of the model (making use of the ROC curve, which allows to evaluate the predictive power of the model without any assumptions on error costs, and possibly R2 from regressions), 5) misclassification analysis could suggest useful insights into determining where the model could be unreliable, 6) implementation stage. The proposed framework has been applied to three applications on different domains, with a view to improve previous research studies. The first developed model predicts mortality within 28 days of patients suffering from acute alcoholic hepatitis. The aim of this application is to build a new predictive model that can be used in clinical practice to identify patients at greatest risk of mortality in 28 days as they may benefit from aggressive intervention, and to monitor their progress while in hospital. A comparison generated by state of the art tools shows an improved predictive power, demonstrating how an appropriate variables inclusion may result in an overall better accuracy of the model, which increased by 25% following an appropriate variables selection process. The second proposed predictive model is designed to aid the diagnosis of dementia, as clinicians often experience difficulties in the diagnosis of dementia due to the intrinsic complexity of the process and lack of comprehensive diagnostic tools. The aim of this application is to improve on the performance of a recent application of Bayesian belief networks using an alternative approach based on logistic regression. The approach based on statistical variables selection outperformed the model which used variables selected by domain experts in previous studies. Obtained results outperform considered benchmarks by 15%. The third built model predicts the probability of experiencing a certain symptom among common side-effects in patients receiving chemotherapy. The newly developed model includes a pre-modelling stage (which was based on previous research studies) and a subsequent regression. The computed accuracy of results (computed on a daily basis for each cycle of therapy) shows that the newly proposed approach has increased its predictive power by 19% when compared to the previously developed model: this has been obtained by an appropriate usage of available a priori knowledge to pre-model the functional forms. As shown by the proposed applications, different aspects of CDSS development are subject to substantial improvements: the application of the proposed framework to different domains leads to more accurate models than the existing state-of-the-art proposals. The developed framework is capable of helping researchers to identify and overcome possible pitfalls in their ongoing research works, by providing them with best practices for each step of the development process. An impact on the development of future clinical decision support systems is envisaged: the usage of an appropriate procedure in model development will produce more reliable and accurate systems, and will have a positive impact on the newly produced medical knowledge which may eventually be included in standard clinical practice.

This thesis investigates perinatal trauma and perinatal mental health, including obsessive compulsive, post-traumatic stress, panic, social phobia, agoraphobia, general anxiety, major depression and postnatal depression symptoms within attachment theory's perspective. It aims to give insight into both caregiving and caretaking experiences of mothers in the pursuit of understanding the aftermath of perinatal trauma Thus it aims to understand first of all, interrelated factors like attachment styles, social support and parental rearing experience in predicting perinatal mental health including anxiety specific symptoms. Then it examines the mediational relationship between support and attachment styles and draws attention to understanding the importance of this relationship in relation to practical implications. This thesis also aims to understand the differences and similarities in various trauma experiences. The final aim of this thesis focuses on the experience of perinatal trauma and the relationship between mothers who experienced previous perinatal trauma and the subsequent infant. The thesis employs both qualitative and quantitative design and analysis techniques.

There is a lack of published literature describing resource and support needs of paediatric prescribers and pharmacists. In order to understand how to support this group of healthcare professionals it is first necessary to identify their current use of resources when prescribing and providing pharmacy services in paediatrics. The methods used in this thesis were mixed. They included: focus groups with prescribers, self-completed questionnaires with paediatric pharmacy staff and paediatric prescribers, interviews with electronic prescribing leaders and documentary analysis of board meeting minutes from paediatric hospitals in England. The resource reported to be used most frequently and most useful by both pharmacists and prescribers was the British National Formulary for Children. The BNFc was reported to be useful due to its current information and ease of use. Pharmacist and prescriber participants reported using a wide range of resources suggesting that there is no single resource that meets their information needs when working in paediatrics. There was general agreement that the current poor availability of some paediatric prescribing information could have an adverse effect on the care of patients. Pharmacy staff reported that an electronic medicines management system improved the supply of medication to inpatients, but described a need for additional development of the system for it to be suitable for all medication supply. Paediatric hospital board minutes reported a range of interventions to improve prescribing, but few reported outcomes. To conclude: this thesis describes the extensive resource needs of both paediatric pharmacists and prescribers. The choice of resource is not affected by the status of its accreditation with NICE, raising a question of the value of this accreditation process. The lack of collaboration between paediatric hospitals regarding strategies used to improve paediatric prescribing is not acceptable and may lead to duplication of work or investment in poor support solutions.

Healthcare systems require continuous monitoring of risk to prevent adverse events. Risk analysis is a time consuming activity that depends on the background of analysts and available data. Patient safety data is often incomplete and biased. This research proposes systematic approaches to monitor risk in healthcare using available patient safety data. The methodologies combine traditional healthcare risk analysis methods with safety theory concepts, in an innovative manner, to allocate available evidence to potential risk sources throughout the system. We propose the use of data mining to analyze near-miss reports and guide the identification of risk sources. In addition, we propose a Maximum-Entropy based approach to monitor risk sources and prioritize investigation efforts accordingly. The products of this research are intended to facilitate risk analysis and allow for timely identification of risks to prevent harm to patients.

Online health communities (OHCs) have become a medium for patients to share their personal experiences and interact with peers on topics related to a disease, medication, side effects, and therapeutic processes. Many studies show that using OHCs regularly decreases mortality and improves patients mental health. As a result of their benefits, OHCs are a popular place for patients to refer to, especially patients with a severe disease, and to receive emotional and informational support. The main reasons for developing OHCs are to present valid and high-quality information and to understand the mechanism of social support in changing patients' mental health. Given the purpose of OHC moderators for developing OHCs applications and the purpose of patients for using OHCs, there is no facility, feature, or sub-application in OHCs to satisfy patient and moderator goals. OHCs are only equipped with a primary search engine that is a keyword-based search tool. In other words, if a patient wants to obtain information about a side-effect, he/she needs to browse many threads in the hope that he/she can find several related comments. In the same way, OHC moderators cannot browse all information which is exchanged among patients to validate their accuracy. Thus, it is critical for OHCs to be equipped with computational tools which are supported by several sophisticated computational models that provide moderators and patients with the collection of messages that they need for making decisions or predictions. We present multiple computational models to alleviate the problem of OHCs in providing specific types of messages in response to the specific moderator and patient needs. Specifically, we focused on proposing computational models for the following tasks: identifying emotional support, which presents OHCs moderators, psychologists, and sociologists with insightful views on the emotional states of individuals and groups, and identifying informational support, which provides patients with an efficient and effective tool for accessing the best-fit messages from a huge amount of patient posts to satisfy their information needs, as well as provides OHC moderators, health-practitioners, nurses, and doctors with an insightful view about the current discussion under the topics of side-effects and therapeutic processes, giving them an opportunity to monitor and validate the exchange of information in OHCs. We proposed hybrid models that combine high-level, abstract features extracted from convolutional neural networks with lexicon-based features and features extracted from long short-term memory networks to capture the semantics of the data. We show that our models, with and without lexicon-based features, outperform strong baselines.

BACKGROUND: Multimorbidity (i.e. the presence of more than one chronic disease in the same patient) and comorbidity (i.e. the presence of more than one chronic disease in the presence of an index disease) are main sources of dysfunction in chronic patients and avoidable costs in conventional health systems worldwide. By affecting a majority of elderly population worldwide, multimorbidity prompts the need for revisiting the single disease approach followed by contemporary clinical practice and elaborate strategies that target shared mechanisms of associated diseases with the potential of preventing, decelerating or even halting multimorbid disease progression. However, our current understanding on disease interactions is rather limited, and although many disorders have been associated based on their shared molecular traits and their observed co-occurrence in different populations, no comprehensive approach has been outlined to translate this knowledge into clinical practice. The advent of novel measurement technologies (e.g. omics) and recent initiatives on digital health (e.g. registries, electronic health records) are facilitating access to an enormous amount of patient-related information from whole populations to molecular levels. State-of-the art computational models and machine learning tools demonstrate high potential for health prediction and together with systems biology are shaping the practicalities of systems medicine. Given the extremely long and expensive bench to clinics cycles of the biomedical sector, systems medicine promises a fast track approach where scientific evidence support clinical care, while simultaneously collected insights from daily clinical practice promote new scientific discoveries and optimize healthcare. The PhD thesis aims to explore multimorbidity from a systems medicine perspective on the concrete and practical use case of chronic obstructive pulmonary disease (COPD). COPD constitutes an ideal use case due to several factors, including: i) its high impact on healthcare and its ever-increasing burden; ii) its heterogeneous disease manifestations, and progress, often involving extra-pulmonary effects, including highly prevalent comorbidities (e.g. type 2 diabetes mellitus, cardiovascular disorders, anxiety-depression and lung cancer); and, iii) its well described systemic effects that are suggested associations with comorbidities in terms of underlying mechanisms. HYPOTHESIS: The central hypothesis of the PhD thesis builds on the emerging biological evidence that clustering of comorbid conditions, a phenomenon seen in complex chronic patients, could be due to shared abnormalities in relevant biological pathways (i.e. bioenergetics, inflammation and tissue remodelling). It is assumed that a systems understanding of the patient conditions may help to uncover the molecular mechanisms and lead to the design of preventive and targeted therapeutic strategies aiming at modulating patient prognosis. The PhD thesis focuses on non-pulmonary phenomena of COPD; that is, systemic effects and comorbidities, often observed in patients with COPD as a paradigm of complex chronic disease. OBJECTIVES: The general objective of the PhD thesis is threefold: i) to investigate molecular disturbances at body systems level that may lead to a better understanding of characteristic systemic effects and comorbidities of patients with COPD; ii) to analyse population level patterns of COPD comorbidities and investigate their role in the health risk of patients with COPD; and, iii) to explore technological strategies and tools that facilitate the transfer of the collected knowledge on comorbidity into clinical practice. MAIN FINDINGS: Firstly, the PhD thesis introduced a novel knowledge management tool for targeted molecular analysis of underlying disease mechanisms of skeletal muscle dysfunction in patients with COPD. Second, a network analysis approach was outlined to further study this systemic effect, as well as the causes of abnormal adaptation of COPD muscle to exercise training. Furthermore, this work together with three other studies also aimed to reveal the general underlying causes of comorbidity clustering in COPD, using different modelling approaches. Overarching outcome of these studies indicates abnormalities in the complex co-regulation of core biological pathways (i.e. bioenergetics, inflammation, oxidative stress and tissue remodelling) both on muscle and body systems level (blood, lung), which paves the way for the development of novel pharmacological and non-pharmacological preventive interventions on non- pulmonary phenomena in patients with COPD. Furthermore, results indicated strong relation of muscle related dysregulations to aerobic capacity, in opposed to pulmonary severity of COPD. These findings have far reaching potential in COPD care, starting from defining the need for better characterization of exercise performance in the clinic practice and the promotion of physical activity from early stages of the disease. This PhD thesis also generated outcomes with respect to the risk of multimorbidity in patients with COPD using a population health approach. The thesis validated that patients with COPD are in increased risk to co-occur with other diseases compared to non-COPD patients, regardless of the population and healthcare system specificities of different regions (i.e. Catalonia, US). These findings indicated the potential role of multimorbidity as a risk factor for COPD, that was evaluated in the PhD thesis by constructing health risk assessment models to predict unexpected medical events in patients with COPD. The promising performance of the models and the prominent role of multimorbidity in these models presented a powerful argument for its role in clinical staging of the disease and their potential in clinical decision support. CONCLUSIONS: The PhD thesis achieved main points of the general objectives, namely: i) to perform a systems analysis of patients with COPD by investigating molecular disturbances at body systems level leading to a better understanding of characteristic systemic effects and comorbidities of patient with COPD; ii) to analyse population level patterns of COPD comorbidities and investigate their role in the health risk of patients with COPD; and iii) to explore technological strategies and tools that facilitate the transfer of the collected knowledge on comorbidity into clinical practice. Accordingly, the following conclusions arise: 1. Non-pulmonary manifestations in patients with Chronic Obstructive Pulmonary Disease (COPD) have a major negative impact on: highly relevant clinical events, use of healthcare resources and prognosis. Accordingly, the following indications were made: a. Actionable insights on non-pulmonary phenomena should be included in the clinical staging of these patients in an operational manner. b. Management of patients with COPD should be revisited to incorporate an integrative approach to non-pulmonary phenomena. c. Innovative cost-effective interventions, and pharmacological and non- pharmacological treatments targeting prevention of non-pulmonary manifestations in patients with COPD should be developed, and properly assessed. 2. Abnormal co-regulation of core biological pathways (i.e. bioenergetics, inflammation, tissue remodelling and oxidative stress), both in skeletal muscle and at body systems level, are common characteristics of patients with COPD, which potentially play a major role in comorbidity clustering. 3. Consistent relationships between cardiovascular health, skeletal muscle dysfunction and clinical outcomes in patients with COPD was identified, which makes it a priority to characterize patient exercise performance and physical activity in the clinic, and to adopt early cardiopulmonary rehabilitation strategies to modulate prognosis and prevent comorbidity clustering in these patients. 4. Multimorbidity is a strong predictor of unplanned medical events in patients with COPD and shows high potential to be used for personalized health risk prediction and service workflow selection. 5. Personalized health risk prediction was identified as a high potential tool for the integration and transfer of scientific evidence on multimorbidity to daily clinical practice. Limiting factors of its present applicability were explored and implementation strategies based on cloud computing solutions were proposed.
INTRODUCCIÓ: Tant la multimorbiditat (la presència de més d'una malaltia crònica en el mateix pacient), com la comorbiditat (la presència de més d'una malaltia crònica quan hi ha una malaltia de referència) són una font important de disfuncions en l’atenció sanitària dels pacients crònics i generen importants despeses evitables en sistemes de salut arreu del món. La multimorbiditat/comorbiditat afecta la majoria de població de més de 65 anys. El seu gran impacte sanitari i social fa necessària la revisió d’aspectes essencials de la pràctica mèdica convencional, molt enfocada al tractament de cada malaltia de forma aïllada. En aquest sentit, cal elaborar estratègies que considerin els mecanismes biològics comuns entre patologies, per tal de prevenir, retardar o fins i tot aturar la progressió del fenomen. Malauradament, el poc coneixement dels mecanismes biològics que modulen les interaccions entre malalties és un factor limitant important. Hi ha estudis sobre els mecanismes moleculars comuns entre malalties i s’han realitzat anàlisis poblacionals de la multimorbiditat, però no existeix encara una aproximació holística per tal de traduir aquest coneixement a la pràctica clínica. L’aparició de noves tecnologies òmiques, així com iniciatives recents en l’àmbit de la salut digital, han facilitat l'accés a una quantitat enorme d'informació dels pacients, tant a nivell poblacional com a nivell molecular. A més, les eines computacionals i d'aprenentatge automàtic existents estan demostrant un gran potencial predictiu que, conjuntament amb les metodologies de la biologia de sistemes, estan conformant els aspectes pràctics del desplegament de la medicina de sistemes. De forma progressiva, aquesta última esdevé una via efectiva per accelerar el rol de l’evidència científica com a suport a la atenció clínica. De forma recíproca, la digitalització sistemàtica de la pràctica clínica diària, permet la generació de noves descobertes científiques i la optimització de l’assistència sanitària. Aquesta tesis doctoral pretén explorar la multimorbiditat des d’una perspectiva de medicina de sistemes, considerant com a cas d'ús concret i pràctic la malaltia pulmonar obstructiva crònica (MPOC). La MPOC constitueix un cas d'ús ideal a causa de diversos factors: i) el seu alt impacte a nivell sanitari; ii) la heterogeneïtat en quant a manifestacions i progrés, sovint amb efectes extra-pulmonars, incloent de forma freqüent comorbiditats com la diabetis mellitus tipus 2, trastorns cardiovasculars, l'ansietat-depressió i el càncer de pulmó; i, iii) els efectes sistèmics de la malaltia pulmonar, que podrien presentar mecanismes biològics comuns a algunes comorbiditats. HIPÒTESIS: La hipòtesi central d’aquesta tesis doctoral considera que la multimorbiditat podria explicar-se per alteracions en les xarxes de regulació de mecanismes biològics rellevants com la bioenergètica, inflamació i remodelació de teixits. En aquest sentit, l’anàlisi holística del problema podria millorar la comprensió dels mecanismes moleculars que modulen les associacions entre malalties i, per tant, facilitar el disseny d'estratègies terapèutiques preventives i dirigides a modular el pronòstic dels pacients. Aquesta tesis doctoral estudia els fenòmens extra-pulmonars de la MPOC; és a dir, efectes sistèmics (disfunció del múscul esquelètic) i comorbiditats, com a paradigma de malalties cròniques complexes. OBJECTIUS: L'objectiu general d’aquesta tesis doctoral és triple: i) l’anàlisi holístic de pacients amb MPOC amb focus en la disfunció muscular i les comorbiditats; ii) avaluar el paper de les comorbiditats en el risc de salut dels pacients amb MPOC, tant a nivell poblacional com individual; i, iii) explorar estratègies tecnològiques i eines de salut digital que facilitin la transferència de coneixement a la pràctica clínica diària. RESULTATS: El primer manuscrit de la tesi descriu una nova eina de gestió del coneixement per l’anàlisi molecular dels mecanismes de disfunció del múscul esquelètic en pacients amb MPOC. També dins el primer objectiu de la tesi, s’efectua un anàlisi de xarxes orientat a la identificació de mòduls biològics explicatius de la disfunció muscular i de l’adaptació anòmala d’aquests malalts a l’entrenament físic, tal com es descriu en el segon manuscrit. Els tres articles següents exploren, des de diferents perspectives, l’impacte i mecanismes de les comorbiditats en els pacients amb MPOC. Els principals resultats d'aquests estudis indiquen una complexa i anormal regulació de vies biològiques principals, com es el cas de la bioenergètica, inflamació, estrès oxidatiu i remodelació de teixits, tant a nivell del múscul com a nivell sistèmic (sang, pulmó). Aquests resultats obren noves vies per a intervencions preventives, tant farmacològiques com no farmacològiques, sobre els fenòmens no pulmonars que presenten els pacients amb MPOC. Els resultats indiquen una associació de les alteracions musculars amb la capacitat aeròbica, i no pas amb la gravetat de la malaltia pulmonar. Aquestes troballes tenen un gran potencial en la millora de la gestió dels pacients amb MPOC, començant per la necessitat d’una millor caracterització de la capacitat aeròbica en la pràctica clínica i la promoció d'activitat física des de les primeres etapes de la malaltia. La tesi també ha generat resultats d’interès en relació amb el risc de multimorbiditat en pacients amb MPOC, mitjançant un enfocament de salut poblacional. Els resultats evidencien que els pacients amb MPOC presenten un risc mes elevat de comorbiditat que els pacients sense MPOC, independentment de les especificitats de la població i del sistema sanitari de les àrees analitzades (Catalunya, EUA). La tesi també demostra el paper de la multimorbiditat com a factor modulador del risc clínic dels pacients amb MPOC. Aquests resultats indiquen l’interès de l’ús de la multimobiditat en l’estadiatge dels pacients amb MPOC i en l’elaboració d’eines de suport al procés de decisió clínica. CONCLUSIONS: Aquesta tesi doctoral ha assolit els objectius generals plantejats i proposa les següents conclusions: 1. Les manifestacions no pulmonars en els pacients amb malaltia pulmonar obstructiva crònica (MPOC) tenen un impacte negatiu respecte a esdeveniments de gran rellevància clínica, ús de recursos sanitaris i pronòstic. En conseqüència, es fan les següents recomanacions: a. Els fenòmens no pulmonars de la MPOC s’haurien d’incloure de manera operativa en l’estadiatge d'aquests pacients. b. S’hauria de redefinir la gestió clínica dels pacients amb MPOC tot incorporant un enfocament holístic dels fenòmens no pulmonars. c. S’haurien de desenvolupar i avaluar correctament noves intervencions, farmacològiques i no farmacològiques, per a la prevenció de les manifestacions no pulmonars en pacients amb MPOC. 2. Les alteracions de la regulació de vies biològiques rellevants com la bioenergètica, inflamació, estrès oxidatiu i la remodelació de teixits a nivell del múscul esquelètic, i també a nivell sistèmic, s’observa en els pacients amb MPOC i pot tenir un paper important en les co-morbiditats. 3. Les relacions entre alteracions cardiovasculars, disfunció del múscul esquelètic i altres aspectes clínics dels pacients amb MPOC, indiquen la necessitat de caracteritzar la capacitat aeròbica i els nivells d'activitat física en la pràctica clínica, així com la implementació d’estratègies de rehabilitació cardiopulmonar en les primeres etapes de la malaltia, per tal de modular la prognosis dels malalts i prevenir l’aparició de comorbiditats. 4. La multimorbiditat és un bon predictor d’esdeveniments clínics rellevants en pacients amb MPOC i mostra un gran potencial per a personalitzar l’estimació de risc i la selecció de serveis. 5. La predicció de risc de forma personalitzada s’ha identificat com una eina amb molt potencial per a la gestió de la multimorbiditat en la pràctica clínica diària. S’han explorat els factors limitants de la seva aplicabilitat i s’han proposat estratègies d'implementació d’eines predictives adients, basades en solucions de computació en el núvol.
INTRODUCCIÓN: Tanto la multimorbilidad (la presencia de más de una enfermedad crónica en un mismo paciente) como la comorbilidad (la presencia de más de una enfermedad crónica en presencia de una enfermedad de referencia) son una fuente importante de disfunciones en la atención sanitaria de los pacientes crónicos y generan importantes costes evitables en los sistemas de salud de todo el mundo. La multimorbilidad/comorbilidad afecta a la mayoría de la población de más de 65 años. Debido a su gran impacto sanitario y social, resulta necesaria la revisión de aspectos esenciales de la práctica médica convencional, muy enfocada en el tratamiento de cada enfermedad de forma aislada. En este sentido, es necesario elaborar estrategias que consideren mecanismos biológicos comunes entre patologías, con el fin de prevenir, retrasar o incluso detener la progresión del fenómeno. Desgraciadamente, el escaso conocimiento de los mecanismos biológicos que modulan las interacciones entre enfermedades es un factor limitante importante. Existen estudios sobre los mecanismos moleculares comunes entre enfermedades y se han realizados análisis poblaciones de la multimorbilidad, pero no existe aún una aproximación holística que permita traducir este conocimiento a la práctica clínica. La aparición de nuevas tecnologías ómicas, así como recientes iniciativas en el ámbito de la salud digital, han facilitado el acceso a una cantidad enorme de información sobre los pacientes, tanto a nivel poblacional como a nivel molecular. Además, las herramientas computacionales y de aprendizaje automático existentes demuestran un gran potencial predictivo que, conjuntamente con las metodologías de biología de sistemas, están conformando los aspectos prácticos de la medicina de sistemas. De manera progresiva esta última se está convirtiendo en una vía efectiva para acelerar el papel de la evidencia científica como soporte a la atención clínica. De forma recíproca, la digitalización sistemática de la práctica clínica diaria permite la generación de nuevos descubrimientos científicos y la optimización de la asistencia sanitaria. Esta tesis doctoral pretende explorar la multimorbilidad desde una perspectiva de medicina de sistemas, considerando como caso de uso concreto y práctico la enfermedad pulmonar obstructiva crónica (EPOC). La EPOC constituye un caso de uso ideal debido a diversos factores: i) su alto impacto a nivel sanitario; ii) la heterogeneidad en cuanto a manifestaciones y progreso, a menudo con efectos extra pulmonares, incluyendo de forma frecuente comorbilidades como la diabetes mellitus tipo 2, trastornos cardiovasculares, la ansiedad-depresión y el cáncer de pulmón; y, iii) los efectos sistémicos de la enfermedad pulmonar, que podrían presentar mecanismos biológicos comunes a algunas comorbilidades. HIPÓTESIS: La hipótesis central de esta tesis doctoral considera que la multimorbilidad podría explicarse por alteraciones en las redes de regulación de mecanismos biológicos relevantes como la bioenergética, inflamación y remodelación de tejidos. En este sentido, el análisis holístico del problema podría mejorar la comprensión de los mecanismos moleculares que modulan las asociaciones entre enfermedades y, por tanto, facilitar el diseño de estrategias terapéuticas preventivas y dirigidas a modular el pronóstico de los pacientes. Esta tesis doctoral estudia los fenómenos extra pulmonares de la EPOC; es decir, efectos sistémicos (disfunción del músculo esquelético) y comorbilidades, como paradigma de enfermedades crónicas complejas. OBJETIVOS: El objetivo general de esta tesis doctoral es triple: i) el análisis holístico de pacientes con EPOC focalizando en la disfunción muscular y la comorbilidades; ii) evaluar el papel de las comorbilidades en el riesgo de salud de los pacientes con EPOC, tanto a nivel poblacional como individual; y, iii) explorar estrategias tecnológicas y herramientas de salud digital que faciliten la transferencia de conocimiento a la práctica clínica diaria. RESULTADOS: El primer manuscrito de la tesis describe una nueva herramienta de gestión del conocimiento para el análisis molecular de los mecanismos de disfunción del músculo esquelético en pacientes con EPOC. Incluido en el primer objetivo de la tesis, se efectúa un análisis de redes orientado a la identificación de módulos biológicos que explican la disfunción muscular y la adaptación anómala de estos pacientes al entrenamiento físico, tal y cómo se describe en el segundo manuscrito. Los tres artículos siguientes exploran, desde perspectivas diferentes, el impacto y mecanismos de las comorbilidades en los pacientes con EPOC. Los principales resultados de estos estudios indican una compleja y anormal regulación de vías biológicas principales, como es el caso de la bioenergética, inflamación, estrés oxidativo y remodelación de tejidos, tanto a nivel del músculo como a nivel sistémico (sangre, pulmón). Estos resultados abren nuevas vías para intervenciones preventivas, tanto farmacológicas como no farmacológicas, sobre los fenómenos no pulmonares que presentan los pacientes con EPOC. Los resultados indican una asociación de las alteraciones musculares con la capacidad aeróbica, y no con la gravedad de la enfermedad pulmonar. Estos hallazgos tienen un gran potencial en la mejora de la gestión de los pacientes con EPOC, empezando por la necesidad de una mejor caracterización de la capacidad aeróbica en la práctica clínica y la promoción de actividad física desde etapas tempranas de la enfermedad. La tesis también ha generado resultados de interés en relación con el riesgo de multimorbilidad en pacientes con EPOC, mediante un enfoque de salud poblacional. Los resultados evidencian que los pacientes con EPOC presentan un mayor riesgo de comorbilidad que los pacientes sin EPOC, independientemente de las especificidades de la población y del sistema sanitario de las áreas analizadas (Cataluña, EUA). La tesis demuestra también el papel de la multimorbilidad como factor modulador del riesgo clínico de los pacientes con EPOC. Estos resultados indican la conveniencia del uso de la multimorbilidad en el estadiaje de los pacientes con EPOC y en la elaboración de herramientas de soporte al proceso de decisión clínica. CONCLUSIONES: Esta tesis doctoral ha conseguido los objetivos generales planteados y propone las siguientes conclusiones: 1. Las manifestaciones no pulmonares en los pacientes con enfermedad pulmonar obstructiva crónica (EPOC) tienen un impacto negativo respecto a eventos de gran relevancia clínica, uso de recursos sanitarios y pronóstico. En consecuencia, se formulan las siguientes recomendaciones: a) Los fenómenos no pulmonares de la EPOC deberían incluirse de manera operativa en el estadiaje de estos pacientes. b) Se debería redefinir la gestión clínica de los pacientes con EPOC incorporando un enfoque holístico de los fenómenos no pulmonares. c) Se deberían desarrollar y evaluar correctamente nuevas intervenciones, farmacológicas y no farmacológicas, para la prevención de las manifestaciones no pulmonares en pacientes con EPOC. 2. Las alteraciones de la regulación de vías biológicas relevantes como la bioenergética, inflamación, estrés oxidativo y la remodelación de tejidos a nivel del músculo esquelético y también a nivel sistémico, se observa en pacientes con EPOC y puede tener un papel importante en las comorbilidades. 3. Las relaciones entre alteraciones cardiovasculares, disfunción del músculo esquelético y otros aspectos clínicos de los pacientes con EPOC, indican la necesidad de caracterizar la capacidad aeróbica y los niveles de actividad física en la práctica clínica, así como la implementación de estrategias de rehabilitación cardiopulmonar en las primeras etapas de la enfermedad, con el fin de modular el pronóstico de los pacientes y prevenir la aparición de comorbilidades. 4. La multimorbilidad es un buen predictor de eventos clínicos relevantes en pacientes con EPOC y muestra un gran potencial para personalizar la estimación de riesgo y la selección de servicios. 5. La predicción del riesgo de forma personalizada se ha identificado como una herramienta con alto potencial para la gestión de la multimorbilidad en la práctica clínica diaria. Se han explorado los factores limitantes de su aplicabilidad y se han propuesto estrategias de implementación de herramientas predictivas adecuadas, basadas en soluciones de computación en la nube.

Clinical diagnosis of chronic disease is a vital and challenging research problem which requires intensive clinical practice guidelines in order to ensure consistent and efficient patient care. Conventional medical diagnosis systems inculcate certain limitations, like complex diagnosis processes, lack of expertise, lack of well described procedures for conducting diagnoses, low computing skills, and so on. Automated clinical decision support system (CDSS) can help physicians and radiologists to overcome these challenges by combining the competency of radiologists and physicians with the capabilities of computers. CDSS depend on many techniques from the fields of image acquisition, image processing, pattern recognition, machine learning as well as optimization for medical data analysis to produce efficient diagnoses. In this dissertation, we discuss the current challenges in designing an efficient CDSS as well as a number of the latest techniques (while identifying best practices for each stage of the framework) to meet these challenges by finding informative patterns in the medical dataset, analysing them and building a descriptive model of the object of interest and thus aiding in medical diagnosis. To meet these challenges, we propose an extension of conventional clinical decision support system framework, by incorporating artificial immune network (AIN) based hyper-parameter optimization as integral part of it. We applied the conventional as well as optimized CDSS on four case studies (most of them comprise medical images) for efficient medical diagnosis and compared the results. The first key contribution is the novel application of a local energy-based shape histogram (LESH) as the feature set for the recognition of abnormalities in mammograms. We investigated the implication of this technique for the mammogram datasets of the Mammographic Image Analysis Society and INbreast. In the evaluation, regions of interest were extracted from the mammograms, their LESH features were calculated, and they were fed to support vector machine (SVM) and echo state network (ESN) classifiers. In addition, the impact of selecting a subset of LESH features based on the classification performance was also observed and benchmarked against a state-of-the-art wavelet based feature extraction method. The second key contribution is to apply the LESH technique to detect lung cancer. The JSRT Digital Image Database of chest radiographs was selected for research experimentation. Prior to LESH feature extraction, we enhanced the radiograph images using a contrast limited adaptive histogram equalization (CLAHE) approach. Selected state-of-the-art cognitive machine learning classifiers, namely the extreme learning machine (ELM), SVM and ESN, were then applied using the LESH extracted features to enable the efficient diagnosis of a correct medical state (the existence of benign or malignant cancer) in the x-ray images. Comparative simulation results, evaluated using the classification accuracy performance measure, were further benchmarked against state-of-the-art wavelet based features, and authenticated the distinct capability of our proposed framework for enhancing the diagnosis outcome. As the third contribution, this thesis presents a novel technique for detecting breast cancer in volumetric medical images based on a three-dimensional (3D) LESH model. It is a hybrid approach, and combines the 3D LESH feature extraction technique with machine learning classifiers to detect breast cancer from MRI images. The proposed system applies CLAHE to the MRI images before extracting the 3D LESH features. Furthermore, a selected subset of features is fed to a machine learning classifier, namely the SVM, ELM or ESN, to detect abnormalities and to distinguish between different stages of abnormality. The results indicate the high performance of the proposed system. When compared with the wavelet-based feature extraction technique, statistical analysis testifies to the significance of our proposed algorithm. The fourth contribution is a novel application of the (AIN) for optimizing machine learning classification algorithms as part of CDSS. We employed our proposed technique in conjunction with selected machine learning classifiers, namely the ELM, SVM and ESN, and validated it using the benchmark medical datasets of PIMA India diabetes and BUPA liver disorders, two-dimensional (2D) medical images, namely MIAS and INbreast and JSRT chest radiographs, as well as on the three-dimensional TCGA-BRCA breast MRI dataset. The results were investigated using the classification accuracy measure and the learning time. We also compared our methodology with the benchmarked multi-objective genetic algorithm (ES)-based optimization technique. The results authenticate the potential of the AIN optimised CDSS.

The utilization of mathematical tools within biology and medicine has traditionally been less widespread compared to other hard sciences, such as physics and chemistry. However, an increased need for tools such as data processing, bioinformatics, statistics, and mathematical modeling, have emerged due to advancements during the last decades. These advancements are partly due to the development of high-throughput experimental procedures and techniques, which produce ever increasing amounts of data. For all aspects of biology and medicine, these data reveal a high level of inter-connectivity between components, which operate on many levels of control, and with multiple feedbacks both between and within each level of control. However, the availability of these large-scale data is not synonymous to a detailed mechanistic understanding of the underlying system. Rather, a mechanistic understanding is gained first when we construct a hypothesis, and test its predictions experimentally. Identifying interesting predictions that are quantitative in nature, generally requires mathematical modeling. This, in turn, requires that the studied system can be formulated into a mathematical model, such as a series of ordinary differential equations, where different hypotheses can be expressed as precise mathematical expressions that influence the output of the model. Within specific sub-domains of biology, the utilization of mathematical models have had a long tradition, such as the modeling done on electrophysiology by Hodgkin and Huxley in the 1950s. However, it is only in recent years, with the arrival of the field known as systems biology that mathematical modeling has become more commonplace. The somewhat slow adaptation of mathematical modeling in biology is partly due to historical differences in training and terminology, as well as in a lack of awareness of showcases illustrating how modeling can make a difference, or even be required, for a correct analysis of the experimental data. In this work, I provide such showcases by demonstrating the universality and applicability of mathematical modeling and hypothesis testing in three disparate biological systems. In Paper II, we demonstrate how mathematical modeling is necessary for the correct interpretation and analysis of dominant negative inhibition data in insulin signaling in primary human adipocytes. In Paper III, we use modeling to determine transport rates across the nuclear membrane in yeast cells, and we show how this technique is superior to traditional curve-fitting methods. We also demonstrate the issue of population heterogeneity and the need to account for individual differences between cells and the population at large. In Paper IV, we use mathematical modeling to reject three hypotheses concerning the phenomenon of facilitation in pyramidal nerve cells in rats and mice. We also show how one surviving hypothesis can explain all data and adequately describe independent validation data. Finally, in Paper I, we develop a method for model selection and discrimination using parametric bootstrapping and the combination of several different empirical distributions of traditional statistical tests. We show how the empirical log-likelihood ratio test is the best combination of two tests and how this can be used, not only for model selection, but also for model discrimination. In conclusion, mathematical modeling is a valuable tool for analyzing data and testing biological hypotheses, regardless of the underlying biological system. Further development of modeling methods and applications are therefore important since these will in all likelihood play a crucial role in all future aspects of biology and medicine, especially in dealing with the burden of increasing amounts of data that is made available with new experimental techniques.
Användandet av matematiska verktyg har inom biologi och medicin traditionellt sett varit mindre utbredd jämfört med andra ämnen inom naturvetenskapen, såsom fysik och kemi. Ett ökat behov av verktyg som databehandling, bioinformatik, statistik och matematisk modellering har trätt fram tack vare framsteg under de senaste decennierna. Dessa framsteg är delvis ett resultat av utvecklingen av storskaliga datainsamlingstekniker. Inom alla områden av biologi och medicin så har dessa data avslöjat en hög nivå av interkonnektivitet mellan komponenter, verksamma på många kontrollnivåer och med flera återkopplingar både mellan och inom varje nivå av kontroll. Tillgång till storskaliga data är emellertid inte synonymt med en detaljerad mekanistisk förståelse för det underliggande systemet. Snarare uppnås en mekanisk förståelse först när vi bygger en hypotes vars prediktioner vi kan testa experimentellt. Att identifiera intressanta prediktioner som är av kvantitativ natur, kräver generellt sett matematisk modellering. Detta kräver i sin tur att det studerade systemet kan formuleras till en matematisk modell, såsom en serie ordinära differentialekvationer, där olika hypoteser kan uttryckas som precisa matematiska uttryck som påverkar modellens output. Inom vissa delområden av biologin har utnyttjandet av matematiska modeller haft en lång tradition, såsom den modellering gjord inom elektrofysiologi av Hodgkin och Huxley på 1950‑talet. Det är emellertid just på senare år, med ankomsten av fältet systembiologi, som matematisk modellering har blivit ett vanligt inslag. Den något långsamma adapteringen av matematisk modellering inom biologi är bl.a. grundad i historiska skillnader i träning och terminologi, samt brist på medvetenhet om exempel som illustrerar hur modellering kan göra skillnad och faktiskt ofta är ett krav för en korrekt analys av experimentella data. I detta arbete tillhandahåller jag sådana exempel och demonstrerar den matematiska modelleringens och hypotestestningens allmängiltighet och tillämpbarhet i tre olika biologiska system. I Arbete II visar vi hur matematisk modellering är nödvändig för en korrekt tolkning och analys av dominant-negativ-inhiberingsdata vid insulinsignalering i primära humana adipocyter. I Arbete III använder vi modellering för att bestämma transporthastigheter över cellkärnmembranet i jästceller, och vi visar hur denna teknik är överlägsen traditionella kurvpassningsmetoder. Vi demonstrerar också frågan om populationsheterogenitet och behovet av att ta hänsyn till individuella skillnader mellan celler och befolkningen som helhet. I Arbete IV använder vi matematisk modellering för att förkasta tre hypoteser om hur fenomenet facilitering uppstår i pyramidala nervceller hos råttor och möss. Vi visar också hur en överlevande hypotes kan beskriva all data, inklusive oberoende valideringsdata. Slutligen utvecklar vi i Arbete I en metod för modellselektion och modelldiskriminering med hjälp av parametrisk ”bootstrapping” samt kombinationen av olika empiriska fördelningar av traditionella statistiska tester. Vi visar hur det empiriska ”log-likelihood-ratio-testet” är den bästa kombinationen av två tester och hur testet är applicerbart, inte bara för modellselektion, utan också för modelldiskriminering. Sammanfattningsvis är matematisk modellering ett värdefullt verktyg för att analysera data och testa biologiska hypoteser, oavsett underliggande biologiskt system. Vidare utveckling av modelleringsmetoder och tillämpningar är därför viktigt eftersom dessa sannolikt kommer att spela en avgörande roll i framtiden för biologi och medicin, särskilt när det gäller att hantera belastningen från ökande datamängder som blir tillgänglig med nya experimentella tekniker.

Multivariate pattern analysis (MVPA) of fMRI data has been growing in popularity due to its sensitivity to networks of brain activation. It is performed in a predictive modeling framework which is natural for implementing brain state prediction and real-time fMRI applications such as brain computer interfaces. Support vector machines (SVM) have been particularly popular for MVPA owing to their high prediction accuracy even with noisy datasets. Recent work has proposed the use of relevance vector machines (RVM) as an alternative to SVM. RVMs are particularly attractive in time sensitive applications such as real-time fMRI since they tend to perform classification faster than SVMs. Despite the use of both methods in fMRI research, little has been done to compare the performance of these two techniques. This study compares RVM to SVM in terms of time and accuracy to determine which is better suited to real-time applications.

La paralysie faciale affecte négativement la vie professionnelle, sociale et personnelle des patients concernés. La récupération des mimiques faciales dans des conditions normales et symétriques permet à ces patients d'améliorer leurs qualités de vie. La rééducation fonctionnelle est une étape clinique importante pour améliorer les qualités des interventions chirurgicales. Cependant, la rééducation faciale reste actuellement un défi scientifique, technologique et clinique majeur. En particulier, l'approche conventionnelle manque des retours quantitatifs et objectifs pour optimiser les gestes et les exercices associées. L'objectif de cette thèse est de développer et d'évaluer un système d'aide à la rééducation fonctionnelle de la mimique faciale. La thèse a six contributions principales : (1) un nouveau processus de génération de la tête patient-spécifique avec la texture à partir d'un capteur sans contact de type Kinect ; (2) un nouveau processus de prédiction du crâne à partir de la surface de la tête en utilisant la modélisation statistique de la forme ; (3) une nouvelle méthode d'évaluation des mouvements de la mimique faciale en se basant sur les propriétés musculaires ; (4) un nouveau système de jeu sérieux pour la rééducation fonctionnelle de la mimique faciale (5) un nouveau système d'aide à la décision clinique pour le visage et (6) un guide de référence pour le développement de systèmes de simulation médicale en considérant la déformation des tissus mous en temps réel. Cette thèse ouvre de nouvelles perspectives liées aux différents domaines de recherche allant de la vision par ordinateur (génération automatique des modèles patient-spécifique à partir d'un capteur visuel), la modélisation biomécanique, et l'ingénierie des systèmes pour la rééducation fonctionnelle de la mimique faciale
Facial disorders negatively affect professional, social, and personal lives of involved patients.Thus, recovery of facial mimics into normal and symmetrical conditions allows these patients to improve their life qualities. Functional rehabilitation of facial disorders is an important clinical step to improve qualities of surgical interventions and drug therapies. However, facialmimic rehabilitation currently remains a major scientific, technological, and clinical challenge.Especially, conventional rehabilitation processes lack of quantitative and objective biofeedbacks. Moreover, rehabilitation exercises just included long-term and repetitive actions. This makes patients less ambitious for completing their training programs. Besides, numerous modeling methods, interaction devices, and system architectures have been successfully employed in clinical applications, but they have not been successfully applied for facial mimic rehabilitation. Consequently, this thesis was conducted to complement these drawbacks by designing a clinical decision-support system for facial mimic rehabilitation. Especially, patientspecific models and serious games were integrated with the system for providing quantitative and objective bio-feedbacks and training motivations. The thesis has six main contributions: (1) a novel real-time subject-specific head generation & animation systems, (2) a novel head-to-skull prediction process, (3) a muscle-oriented patientspecific facial paralysis grading system, (4) a novel serious game system for facial mimic rehabilitation, (5) a novel clinical decision-support system for facial mimic rehabilitation, and (6) a reference guide for developing real-time soft-tissues simulation systems. This thesis opens new avenues for new research areas relating to automatic generation of patient specific head from visual sensor and internal structures using statistical shape modeling and real-time modeling and simulation for facial mimic rehabilitation

La découverte de nouveaux médicaments est un processus complexe qui implique de gros investissements et de longues phases de développement. Le coût des échecs est phénoménal et le taux de succès diminue depuis deux décennies, suggérant que le processus d'innovation-développement tel qu'il est pratiqué aujourd'hui n'est plus adapté aux défis actuels. Dans ce travail de thèse, nous proposons une nouvelle approche, dite descendante « Top-down » qui consiste à inverser le procédé traditionnel. Elle reposera essentiellement sur la conception et l'optimisation de modèles de prédiction, afin de mieux orienter les phases successives du développement et de prédire les résultats de chaque étape avec le maximum de précision. Cette démarche s'appuie sur la conception du modèle d'innovation thérapeutique et de toutes ses composantes, notamment le modèle physiopathologique (modèle simulant l'organisme et l'évolution naturelle de la maladie) le modèle thérapeutique (modèle simulant l'effet du médicament sur la maladie), et le modèle d'effet (modèle qui peut être prédit à partir des modèles précédents et qui permet de simuler l'impact du médicament sur la population traitée à partir de son effet sur chaque individu de la population. Le processus complet de notre stratégie comprend dix étapes : élaboration du modèle physiopathologique, calibration, validation, identification des cibles potentielles, activation du logiciel de modifications séquentielles des cibles, construction de la population virtuelle (réaliste ou non), obtention du modèle d'effet nécessaire pour calculer le nombre d'événements évités (NEE), calcul du NEE par cible, comparaison et choix de la “meilleure” cible
New drug discovery and development is a complex process which requires massive investments over protracted horizons. The cost of failed programs is significant in absolute dollar terms. Decreasing R&D productivity over the past 2 decades suggests that the traditional innovation model needs a radical rethink. The central purpose of this thesis work is to lay the theoretical foundations for a new "topdown" approach to new target identification construed as an alternative to the current bottomup approach based on high throughput screening. In essence, it will consist in the design and optimization of in silico models in order to guide the development of a new drug candidate by predicting the outcomes of each successive phase of the development process. This in silico framework brings together a physiopathological model (to simulate the natural evolution of the disease) with a therapeutic model (to simulate the effects of a drug candidate on disease evolution) and the effect model (to predict the impact of the drug candidate on a population of patients)

Introduction :La télésurveillance des stimulateurs cardiaques et défibrillateurs sera à terme le standard pour le suivi des patients implantés. Pourtant, des alertes très nombreuses sont générées par ces dispositifs, et constituent un fardeau pour la prise en charge médicale. De plus, les alertes générées le sont indépendamment du contexte médical individuel du patient, et elles pourraient donc être mieux caractérisées. Cette thèse propose un outil de traitement automatique des alertes générées par la survenue de fibrillation atriale, et basé sur une modélisation des connaissances médicales de type ontologie en OWL2. En particulier, le score de risque cardio-embolique CHA2DS2VASc a été évalué par le biais de l’ontologie, ainsi que le statut d’anticoagulation du patient. Matériel et Méthodes :Une ontologie d’application a été créée en OWL2, afin de représenter les concepts nécessaires au raisonnement sur les alertes. Cette ontologie a été utilisée pour raisonner sur 1783 alertes de FA détectées chez 60 porteurs de stimulateurs cardiaques. Les alertes ont été classées automatiquement selon leur importance d’après une échelle de gravité de 1 à 4. La classification automatique a été comparée à celle réalisée par 2 experts médicaux comme référence. Résultats : 1749 alertes sur 1783 (98%) ont été classées correctement. 58 des 60 patients avaient toutes leurs alertes classées à l’identique par le système testé et par les évaluateurs-médecins. Une approche basée ontologie est à même de permettre un raisonnement automatique sur des données issues de dispositifs médicaux connectés, en les contextualisant en fonction des données médicales individuelles du patient
Introduction :Remote monitoring of cardiac implantable electronic devices (CIED) such as pacemakers and defibrillators is the new follow-up standard. However, the numerous alerts generated in remote monitoring causes a burden for physicians. Morever, many alerts are notified despite the knowledge of patient condition and could be refined. This work proposes an automatic tool for classifying atrial fibrillation alert, based on an ontological knowledge model in OWL2. In particular, CHA2DS2VASc thrombo-embolic risk score and patient anticogulation status are accounted in order to determine alert importance. Materials and methods :An application ontology was designed in OWL2, in order to represent the concepts needed for processing alerts. This ontology was used to infer the importance of 1783 AF alerts among 60 CIED recipients, using a 4-grade scale. Automatic classification was compared to that of 2 medical experts.Results :1749 of 1783 alerts (98%) were correctly classified. 58 of 60 patients had every alerts classified with the same importance by the prototype and the human experts. An ontology-driven automatic reasoning tool is able to classify remote monitoring alerts, by using individual medical context. This technology could be important for managing data generated by connected medical devices

This thesis proposes an algorithm that determine the geometry of 3D-printed, custom-designed spring element bands made of thermoplastic elastomer (TPE) for use in a wearable orthotic device to aid in the physical therapy of a human hand exhibiting spasticity after stroke. Each finger of the hand is modeled as a mechanical system consisting of a triple-rod pendulum with nonlinear stiffness at each joint and forces applied at the attachment point of each flexor muscle. The system is assumed quasi-static, which leads to a torque balance between the flexor tendons in the hand, joint stiffness and the design force applied to the fingertip by the 3D-printed spring element. To better understand material properties of the spring element’s material, several tests are performed on TPE specimens printed with different infill geometries, including tensile tests and cyclic loading tests. The data and stress-strain curves for each geometry type are presented, which yield a nonlinear relationship between stress and strain as well as apparent hysteresis. Polynomial curves are used to fit the data, which allows for the band geometry to be designed. A hypothetical hand is presented along with how input measurements might be taken for the algorithm. The inputs are entered into the algorithm, and the geometry of the bands for each finger are generated. Results are discussed, and future work is noted, providing a means for the design of a customized orthotic device.

Introdução: A Medicina Baseada em Evidências, apesar da grande profusão de publicações da área, enfrenta desafios no intuito de melhorar a qualidade da assistência à saúde. O conhecimento gerado por suas publicações demora a ser posta em prática. Os softwares CDSS de apoio à decisão clínica, podem ser a solução de incorporação das evidências na prática clínica. Esses sistemas já foram associados a melhorias na qualidade de diversos aspectos da assistência à saúde, como a organização, minimização de erros, redução de custos, aumento da eficiência dos cuidados, mas pesquisas com desfechos centrados no paciente ainda são raras. Como outra qualquer intervenção em saúde, as afirmações de que os CDSS são benéficos para o paciente necessitam de confirmação por ensaios clínicos. Objetivos: Verificar se o uso dos CDSS com base em evidências, está associado com melhores resultados clínicos orientados para o paciente. Métodos: Revisão sistemática da literatura dos ensaios clínicos controlados e randomizados que compararam diretamente o uso de CDSS com práticas clínicas convencionais considerando os desfechos clínicos classificados como orientados para o paciente. Resultados: Nossa estratégia de pesquisa identificou 51283 artigos na base MEDLINE-PubMed, sendo 311 selecionados para leitura de título e resumo após a aplicação do filtro para ensaio clínico randomizado, 45 selecionados para leitura do texto completo, dos quais 19 preencheram o critério de elegibilidade. Outros 9 ensaios foram incluídos através da realização de um overview das revisões sistemáticas anteriores. Os ensaios foram publicados entre os anos de 1995 e 2015 e realizados em cinco contextos assistenciais, com duração máxima de 12 meses. A maioria das fontes de evidências que alimentaram os sistemas foram diretrizes de órgão governamental ou sociedades de especialidades. Doze ensaios avaliaram mortalidade, 14 avaliaram hospitalizações ou atendimento de emergência e 6 avaliaram desfechos relacionados a presença de sintomas. Foram realizadas meta-análises de acordo com o contexto assistencial e o tipo de desfecho. Somente uma meta-análise envolvendo a mortalidade de pacientes tratados em ambulatório por diferentes condições clínicas se mostrou estatisticamente significante, favorável ao grupo CDSS, em 3 ensaios randomizados por aglomerado, com risco de viés considerado moderado, que compromete a qualidade da evidência. Conclusões Apesar do potencial dos CDSS no apoio de intervenções de saúde, não há evidência de boa qualidade de que sejam efetivos para aumentar a sobrevida ou a qualidade de vida dos pacientes. O número de ensaios que avaliam esses desfechos, os períodos de tempo pelos quais os pacientes foram seguidos, o número insuficiente de participantes, bem como a heterogeneidade entre os estudos analisados quanto aos cenários clínicos e as fontes de informação que alimentam os softwares não permitiram resultados mais conclusivos
Background: In spite of the wealth of publications in the field, Evidence-Based Medicine faces challenges in order to improve quality of health care. It takes too long for knowledge produced by its publications to be put into practice. Clinical Decision Support Systems (CDSS) may be a solution for incorporation of evidence into clinical practice. These systems have been associated with improvements in quality of various aspects of health care, including its organization, error minimizations, cost reductions and increases in its efficiency, but patient-oriented outcomes are still rare in research literature. Like any other healthcare intervention, claims that CDSS are beneficial for patients need to be confirmed by clinical trials. Objective: To verify whether the use of evidence-based Clinical Decision Support Systems is associated with improved patient-oriented clinical outcomes. Methods: Systematic literature review of randomized controlled trials that directly compared the use of CDSS with usual practice considering clinical outcomes classified as patient-oriented. Results: Our search strategy has identified 51,283 entries in MEDLINE-PubMed and, after filtering for randomized controlled trials 311 papers were selected for title and abstract reading. Forty-five were selected for full-text reading of which 19 have met eligibility criteria. Another nine trials were included after an overview of previous systematic reviews. Trials were published between 1995 and 2015 and performed in five care settings with a maximum follow-up of 12 months. Most evidence sources feeding systems´ knowledge bases were government agency guidelines or specialty societies. Twelve trials have assessed mortality, 14 have assessed hospital admissions and/or emergency visits and nine have assessed symptom-related outcomes. Meta-analyses were performed according to trials´ care setting and outcome types. Only a meta-analysis of three cluster-randomized trials involving mortality among outpatients with different clinical conditions was statistically significant, favouring CDSS group, but risk of bias was moderate, compromising the quality of evidence. Conclusions: Despite the potential of CDSS to improve healthcare quality there is no reliable evidence that they improve patients´ life extension or quality. The insufficient numbers of trials assessing these outcomes, studies´ subjects and follow-up periods, the heterogeneities of clinical settings across studies and knowledge bases feeding the systems impede achieving results that are more conclusive

Personalized medicine offers the most effective treatment protocols to the individual Chronic Myeloid Leukemia (CML) patients. Understanding the molecular biology that causes CML assists in providing efficient treatment. After the identification of an activated tyrosine kinase BCR-ABL1 as the causative lesion in CML, the first-generation Tyrosine Kinase inhibitors (TKI) imatinib (Glivec®), were developed to inhibit BCR-ABL1 activity and approved as a treatment for CML. Despite the remarkable increase in the survival rate of CML patients treated with imatinib, some patients discontinued imatinib therapy due to intolerance, resistance or progression. These patients may benefit from the use of secondgeneration TKIs, such as nilotinib (Tasigna®) and dasatinib (Sprycel®). All three of these TKIs are currently approved for use as frontline treatments. Prognostic scores and molecularbased predictive assays are used to personalize the care of CML patients by allocating risk groups and predicting responses to therapy. Although prognostic scores remain in use today, they are often inadequate for three main reasons. Firstly, since each prognostic score may generate conflicting prognoses for the risk index and it can be difficult to know how to treat patients with conflicting prognoses. Secondly, since prognostic score systems are developed over time, patients can benefit from newly developed systems and information. Finally, the earlier scores use mostly clinically oriented factors instead of those directly related to genetic or molecular indicators. As the current CML treatment guidelines recommend the use of TKI therapy, a new tool that combines the well-known, molecular-based predictive assays to predict molecular response to TKI has not been considered in previous research. Therefore, the main goal of this research is to improve the ability to manage CML disease in individual CML patients and support CML physicians in TKI therapy treatment selection by correctly allocating patients to risk groups and predicting their molecular response to the selected treatment. To achieve this objective, the research detailed here focuses on developing a prognostic model and a predictive model for use as a personalized medicine support system. The system will be considered a knowledge-based clinical decision support system that includes two models embedded in a decision tree. The main idea is to classify patients into risk groups using the prognostic model, while the patients identified as part of the high-risk group should be considered for more aggressive imatinib therapy or switched to secondgeneration TKI with close monitoring. For patients assigned to the low-risk group to imatinib should be predicted using the predictive model. The outcomes should be evaluated by comparing the results of these models with the actual responses to imatinib in patients from a previous medical trial and from patients admitted to hospitals. Validating such a predictive system could greatly assist clinicians in clinical decision-making geared toward individualized medicine. Our findings suggest that the system provides treatment recommendations that could help improve overall healthcare for CML patients. Study limitations included the impact of diversity on human expertise, changing predictive factors, population and prediction endpoints, the impact of time and patient personal issues. Further intensive research activities based on the development of a new predictive model and the method for selecting predictive factors and validation can be expanded to other health organizations and the development of models to predict responses to other TKIs.
Thesis (Ph.D.) -- University of Adelaide, School of Computer Science, 2018

This dissertation looks at how new sources of information should be incorporated into medical decision-making processes to improve patient outcomes and reduce costs. There are three fundamental challenges that must be overcome to effectively use personalized medicine, we need to understand: 1) how best to appropriately designate which patients will receive the greatest value from these processes; 2) how physicians and caregivers interpret additional patient-specific information and how that affects their decision-making processes; and finally, (3) how to account for a patient’s ability to engage in their own healthcare decisions. The first study looks at how we can infer which patients will receive the most value from genomic testing. The difficult statistical problem is how to separate the distribution of patients, based on ex-ante factors, to identify the best candidates for personalized testing. A model was constructed to infer a healthcare provider’s decision on whether this test would provide beneficial information in selecting a patient’s medication. Model analysis shows that healthcare providers’ primary focus is to maximize patient health outcomes while considering the impact the patient’s economic welfare. The second study focuses on understanding how technology-enabled continuity of care (TECC) for Chronic Obstructive Pulmonary Disease (COPD) and Congestive Heart Failure (CHF) patients can be utilized to improve patient engagement, measured in terms of patient activation. We shed light on the fact that different types of patients garnered different levels of value from the use of TECC. The third study looks at how data-driven decision support systems can allow physicians to more accurately understand which patients are at high-risk of readmission. We look at how we can use available patient-specific information for patients admitted with CHF to more accurately identify which patients are most likely to be readmitted, and also why – whether for condition-related reasons versus for non- related reasons, allowing physicians to suggest different patient-specific readmission prevention strategies. Taken together, these three studies allow us to build a robust theory to tackle these challenges, both operational and policy-related, that need to be addressed for physicians to take advantage of the growing availability of patient-specific information to improve personalized medication processes.

Lung cancer is the leading cause of cancer-related deaths among men and women. Non-Small Cell Lung Cancer (NSCLC) constitutes the most common type of lung cancer and is frequently diagnosed at advanced stages. In the past decade, discovery of Epidermal Growth Factor Receptor (EGFR) mutations have heralded a new paradigm of personalized treatment for NSCLC. Clinical studies have shown that molecular targeted therapies increase survival and improve quality of life in patients. Despite these advances, the realization of personalized therapies for NSCLC faces a number of challenges including the integration of clinical and genetic data and a lack of clinical decision support tools to assist physicians with patient selection. This thesis demonstrates the development of a predictive computational model for personalized therapeutic interventions in advanced NSCLC. The findings suggest that the combination of clinical and genetic data significantly improves the model’s predictive performance for tumor response than clinical data alone.

Mobile devices, such as smartphones and wearable devices, have become a popular platform to deliver recommendations and interact with users. To learn the decision rule of assigning recommendations, i.e. policy, neither one homogeneous policy for all users nor completely heterogeneous policy for each user is appropriate. Many attempts have been made to learn a policy for making recommendations using observational mobile health (mHealth) data. The majority of them focuses on a homogeneous policy, that is a one-fit-to-all policy for all users. It is a fair starting point for mHealth study, but it ignores the underlying user heterogeneity. Users with similar behavior pattern may have unobservable underlying heterogeneity. To solve this problem, we develop a personalized learning framework that models both population and personalized effect simultaneously. In the first part of this dissertation, we address the personalized policy learning problem using longitudinal mHealth application usage data. Personalized policy represents a paradigm shift from developing a single policy that may prescribe personalized decisions by tailoring. Specifically, we aim to develop the best policy, one per user, based on estimating random effects under generalized linear mixed model. With many random effects, we consider new estimation method and penalized objective to circumvent high-dimensional integrals for marginal likelihood approximation. We establish consistency and optimality of our method with endogenous application usage. We apply our method to develop personalized prompt schedules in 294 application users, with a goal to maximize the prompt response rate given past application usage and other contextual factors. We found the best push schedule given the same covariates varied among the users, thus calling for personalized policies. Using the estimated personalized policies would have achieved a mean prompt response rate of 23% in these users at 16 weeks or later: this is a remarkable improvement on the observed rate (11%), while the literature suggests 3%-15% user engagement at 3 months after download. The proposed method compares favorably to existing estimation methods including using the R function glmer in a simulation study. In the second part of this dissertation, we aim to solve a practical problem in the mHealth area. Low response rate has been a major issue that blocks researchers from collecting high quality mHealth data. Therefore, developing a prompting system is important to keep user engagement and increase response rate. We aim to learn personalized prompting time for users in order to gain a high response rate. An extension of the personalized learning algorithm is applied on the Intellicare data that incorporates penalties of the population effect parameters and personalized effect parameters into learning the personalized decision rule of sending prompts. The number of personalized policy parameters increases with sample size. Since there is a large number of users in the Intellicare data, it is challenging to estimate such high dimensional parameters. To solve the computational issue, we employ a bagging method that first bootstraps subsamples and then ensembles parameters learned from each subsample. The analysis of Intellicare data shows that sending prompts at a personalized hour helps achieve a higher response rate compared to a one-fit-to-all prompting hour.

Despite considerable effort considerable cost in both time and money, as many as two out of three persons with type 1 diabetes are not in control of their disease. As a result, 40% of these individuals will go on to develop at least one serious complication including retinopathy, nephropathy, neuropathy and cardiomyopathy. It is further estimated that as much as $4 billion could be saved annually if all persons with type 1 diabetes in the US were properly controlled. Adequate treatment of type 1 diabetes is predicated on the estimation of three clinical insulin therapy parameters: the basal dose, the insulin sensitivity factor and the insulin-to-carbohydrate ratio. Currently, these therapy parameters are determined by iterative titration procedures based on expert opinion. Unfortunately, there is evidence suggesting that for the majority of individuals, these titration protocols do not provide good results. In this work we develop an alternative to traditional insulin titration protocols that allows clinical insulin therapy parameters to be estimated directly from a set of easily acquired measurements. First, a simple model of type 1 diabetes is used to derive a series of equations connecting the model's parameters to the clinically important insulin therapy parameters of insulin sensitivity factor, insulin-to-carbohydrate ratio and basal insulin dose. The simplifying assumptions used to derive these equations are tested and shown to be valid and the Fisher Information Matrix is used to demonstrate parameter identifiability. Parameter estimation is then performed on two cohorts of virtual subjects, as well as two segments of real continuous glucose monitoring data from a person with type 1 diabetes. Identification of the true insulin therapy parameters is successful under most conditions for both cohorts of virtual subjects. Parameter estimation for one of the two segments of real continuous glucose monitoring data is also successful. Finally, because continuous glucose monitors are instrumental to successful implementation of our insulin therapy framework, the physiological environment in which continuous glucose monitoring takes place is modeled and a fundamental limitation on measurement precision is shown to exist. An examination of physiological variability in the parameters indicates that many of the challenges observed in real world continuous glucose monitoring may have a relationship to changes in capillary bed perfusion. A rationale for anecdotally reported sensor faults is also proposed based on the physical mechanisms explored.
text

Personalized cancer medicine is a cancer care paradigm in which diagnostic and therapeutic strategies are customized for individual patients. Microsystems that are created by Micro-Electro-Mechanical Systems (MEMS) technology and integrate various diagnostic and therapeutic methods on a single chip hold great potential to enable personalized cancer medicine. Toward ultimate realization of such microsystems, this thesis focuses on developing critical functional building blocks that perform genetic variation identification (single-nucleotide polymorphism (SNP) genotyping) and specific, efficient and flexible cell manipulation on microfluidic surfaces. For the identification of genetic variations, we first present a bead-based approach to detect single-base mutations by performing single-base extension (SBE) of SNP specific primers on solid surfaces. Successful genotyping of the SNP on exon 1 of HBB gene demonstrates the potential of the device for simple, rapid, and accurate detection of SNPs. In addition, a multi-step solution-based approach, which integrates SBE with mass-tagged dideoxynucleotides and solid-phase purification of extension products, is also presented. Rapid, accurate and simultaneous detection of 4 loci on a synthetic template demonstrates the capability of multiplex genotyping with reduced consumption of samples and reagents. For cell manipulation, we first present a microfluidic device for cell purification with surface-immobilized aptamers, exploiting the strong temperature dependence of the affinity binding between aptamers and cells. Further, we demonstrate the feasibility of using aptamers to specifically separate target cells from a heterogeneous solution and employing environmental changes to retrieve purified cells. Moreover, spatially specific capture and selective temperature-mediated release of cells on design-specified areas is presented, which demonstrates the ability to establish cell arrays on pre-defined regions and to collect only specifically selected cell groups for downstream analysis. We also investigate tunable microfluidic trapping of cells by exploiting the large compliance of elastomers to create an array of cell-trapping microstructures, whose dimensions can be mechanically modulated by inducing uniform strain via the application of external force. Cell trapping under different strain modulations has been studied, and capture of a predetermined number of cells, from single cells to multiple cells, has been achieved. In addition, to address the lack of aptamers for targets of interest, which is a major hindrance to aptamer-based cell manipulation, we present a microfluidic device for synthetically isolating cell-targeting aptamers from a randomized single-strand DNA (ssDNA) library, integrating cell culturing with affinity selection and amplification of cell-binding ssDNA. Multi-round aptamer isolation on a single chip has also been realized by using pressure-driven flow. Finally, some perspectives on future work are presented, and strategies and notable issues are discussed for further development of MEMS/microfluidics-based devices for personalized cancer medicine.

Dissertação de mestrado em Engenharia Informática (área de especialização em Sistemas Inteligentes)
Currently in the health sector there is a growing need to standardize and promote the improvement of clinical practice in order to reduce costs, which requires a solution that will allow these goals to be more easily achieved. To this end, the solution that gathers the current interest is the use of clinical protocols and promoting conformity with practices contained in them. Clinical protocols aim to improve the quality of the clinical process, reducing variations in clinical practice and reducing health care costs. In order to be effective, these parameters must be integrated into the care flow and provide specific advice to a patient, regardless of time or place. Thus, their formalization as Computer-Interpretable Guidelines (CIG) makes possible the development of decision support systems based on CIGs, which may have a greater impact on the behavior of health professionals. However, the absence of a general pattern in terms of CIG often hinders progress in the development of these systems. Currently available tools for creating and editing clinical protocols for automatic interpretation are not functional or user-friendly. Most of them are academic projects developed in obsolete languages. As a means to solve this issue, this dissertation project presents an user-friendly tool that manages the creation and editing of CIGs, without requiring the user to have programming knowledge, and through the use of interfaces that are simple and intuitive.
Atualmente no setor da saúde há uma crescente necessidade de padronizar e promover a melhoria das práticas clínicas com o intuito de reduzir custos, o que exige uma solução que permita que estes objetivos sejam mais facilmente atingidos. Para o efeito, a solução que mais desperta o interesse atualmente é a utilização de protocolos clínicos e reforço da conformidade com as práticas que neles são recomendadas. Os protocolos clínicos visam melhorar a qualidade do processo clínico, reduzindo as variações da prática clínica e reduzindo os custos de saúde. De forma a serem eficazes, devem ser integrados no fluxo de atendimento e prestar aconselhamento específico para um paciente, independentemente do tempo ou local onde se encontram. Assim, a sua formalização como Computer-Interpretable Guidelines (CIGs) torna possível o desenvolvimento de sistemas de apoio à decisão baseados em CIGs, que apresentam uma maior capacidade de afetar o comportamento dos profissionais de saúde. Contudo, a inexistência de um padrão generalizado a nível das CIGs dificulta muitas vezes o progresso no desenvolvimento destes sistemas. As ferramentas atualmente disponíveis para a criação e edição de protocolos clínicos para interpretação automática não são funcionais ou de fácil utilização. Como meio de resolver esta questão, neste projeto de dissertação propõe-se o desenvolvimento de uma ferramenta user-friendly capaz de gerir a criação e edição de CIGs, sem a necessidade do utilizador apresentar conhecimentos de programação, e através do uso de interfaces que sejam simples e intuitivas.

Dissertação de Mestrado apresentada à Faculdade de Medicina da Universidade de Coimbra com vista à obtenção do grau de Mestre no âmbito do ciclo de estudos de Mestrado Integrado em Medicina
Objective: The goal of this paper is to discuss biological rhythms, focusing on chronotherapy in cancer. The objectives are to: (1) briefly describe the circadian timing system, its physiology and networks; (2) address causal issues that have prompt progress toward an understanding of mechanisms underlying diseases as circadian-based disorders, specifically cancer; (3) review the concepts and principles of chronotherapy, applied in the medical oncology area; (4) dissect the results obtained by comparative studies between chronotherapy and conventional scheduled cancer treatments; and (5) offer a perspective about the future of chronotherapy and its knowledge in oncology. Methods: Review, synthesis, and interpretation of the literature. Results: Biological rhythms are a ubiquitous feature of life. There is circadian synchronization of endless molecular, physiological, biochemical and behavioral processes. Any deregulation of those rhythms may lead to disease, namely cancer. Likewise, experimental and clinical cancer processes are accelerated under rhythm disruption. On the other hand, anticancer drugs have their pharmacologic effects modified up to several folds accordingly to administration time: improved efficacy is seen when drugs are given near their respective times of best tolerability. Data extrapolated from animal experiments allowed a chronomodulated approach in cancer and randomized trials comparing chronotherapy versus conventional treatments have been performed. Besides the fact that some particular endpoints didn’t give always preference to circadian-based therapies, in no case to date has chronotherapy been shown to be less effective than standard approaches. Chronomodulated schedules allow an increase in dose intensity and have a better tolerability profile. Importantly, optimal circadian timing and dosing of anticancer drugs can differ according to gender. Conclusions: Understanding the chronobiology principles has the potential to contribute to improve outcomes, and can open research ground for the development of better prevention and treatment strategies. The fundamental principles of chronotherapy are worthy of further clinical implementation and the future advances towards personalized cancer chronotherapeutics.
Esta tese centra-se na discussão dos ritmos biológicos, dando particular enfoque à cronoterapia em oncologia. Os objetivos são: (1) descrever sumariamente o sistema circadiano, a sua fisiologia e interações; (2) explorar os fundamentos que têm permitido encarar a doença enquanto consequência de distúrbios da estrutura temporal, em especial as doenças oncológicas; (3) rever os conceitos e princípios da cronoterapia, através da sua aplicação em oncologia; (4) analisar os dados resultantes de estudos comparativos opondo duas estratégias ao tratamento de neoplasias: a cronoterapia e metodologias convencionais; e (5) apresentar uma perspectiva acerca do futuro da cronoterapia e dos conhecimentos que lhe estão inerentes em oncologia. Métodos: Revisão, síntese e interpretação da literatura. Resultados: Os ritmos biológicos são uma característica fundamental da vida. Inúmeros processos moleculares, fisiológicos, bioquímicos e comportamentais estão sob a alçada da sincronização circadiana. Uma desregulação desses ritmos pode conduzir a processos patológicos, nomeadamente oncológicos. De facto, processos neoplásicos demonstraram estar acelerados na presença de alterações dos ritmos circadianos tanto em contexto clínico como em experimental. Por outro lado, os efeitos farmacológicos dos medicamentos anti-neoplásicos diferem enormemente de acordo com o momento no tempo em que são administrados: observa-se uma melhor eficácia quando a administração coincide com o respetivo período de melhor tolerância. Resultados extrapolados de estudos pré-clínicos permitiram delinear uma abordagem cronomodulada ao cancro e desenharam-se ensaios clínicos confrontando a cronoterapia com as abordagens terapêuticas convencionais. Alguns resultados em particular não deram sempre preferência à terapêutica baseada nos ritmos circadianos, mas até à data a cronoterapia nunca mostrou ser menos eficaz do que as metodologias de tratamento correntes. Protocolos cronomodulados permitem um aumento na intensidade da dose administrada e normalmente apresentam um melhor perfil de tolerância. É de realçar, ainda, que o momento ótimo e a dose a administrar podem diferir de acordo com o género do doente. Conclusões: A compreensão dos princípios da cronobiologia tem o potencial de contribuir para melhorar os resultados obtidos em oncologia e abre portas à pesquisa de aprimoradas estratégias de prevenção e tratamento. Os princípios fundamentais da cronoterapia carecem de mais investigação e de consequente implementação clínica. O futuro está a avançar no sentido da personalização dos tratamentos cronomodulados em oncologia.

Abstract Background Cardiovascular disease (CVD) has an enormous impact on the lives and health of New Zealanders. There is substantial epidemiological evidence that supports identifying people at high risk of CVD and treating them with lifestyle and drug-based interventions. If fully implemented, this targeted high risk approach could reduce future CVD events by over 50%. Recent studies have shown that a formal CVD risk assessment to the systematically identify high risk patients is rarely done in routine New Zealand general practice and audits of CVD risk management have shown large evidence-practice gaps. The CVD risk prediction score recommended by New Zealand guidelines for identifying high CVD risk patients was derived from the US Framingham Heart Study using data collected between the 1960s and 1980s. This score has only modest prediction accuracy and there are particular concerns about it’s validity for New Zealand sub-populations such as high risk ethnic groups or people with diabetes. Aims The overall aims of this thesis were to investigate the potential of a computerised decision support system (CDSS) to improve the assessment and management of CVD risk in New Zealand general practice while simultaneously developing a sustainable cohort study that could be used for validating and improving CVD risk prediction scores and related research. Methods An environmental scan of the New Zealand health care setting’s readiness to support a CDSS was conducted .The epidemiological evidence was reviewed to assess the effect of decision support systems on the quality of health care and the types and functionality of systems most likely to be successful. This was followed by a focused systematic review of randomised trials evaluating the impact of CDSS on CVD risk assessment and management practices and patient CVD outcomes in primary care. A web-based CDSS (PREDICT) was collaboratively developed. This rules-based provider-initiated system with audit and feedback and referral functionalities was fully integrated with general practice electronic medical records in a number of primary health organisations (PHOs). The evidence-based content was derived from national CVD and diabetes guidelines. When clinicians used PREDICT at the time of a consultation, treatment recommendations tailored to the patient’s CVD and diabetes risk profile were delivered to support decision-making within seconds. Simultaneously, the patient’s CVD risk profiles were securely stored on a central server. With PHO permission, anonymised patient data were linked via encrypted patient National Health Index numbers to national death and hospitalisation data. Three analytical studies using these data are described in this thesis. The first evaluated changes in GP risk assessment practice following implementation of PREDICT; the second investigated patterns of use of the CDSS by GPs and practice nurses; and the third describes the emerging PREDICT cohort and a preliminary validation of risk prediction scores. Results Given the rapid development of organised primary care since the 1990’s, the high degree of general practice computerisation and the New Zealand policy (health, informatics, privacy) environment, the introduction of a CDSS into the primary care setting was deemed feasible. The evidence for the impact of CDSS in general has been moderately favourable in terms of improving desired practice. Of the randomised trials of CDSS for assessing or managing CVD risk, about two-thirds reported improvements in provider processes and two-fifths reported some improvements in intermediate patient outcomes. No adverse effects were reported. Since 2002, the PREDICT CDSS has been implemented progressively in PHOs within Northland and the three Auckland regional District Health Board catchments, covering a population of 1.5 million. A before-after audit conducted in three large PHOs showed that CVD risk documentation increased four fold after the implementation of PREDICT. To date, the PREDICT dataset includes around 63,000 risk assessments conducted on a cohort of over 48,000 people by over 1000 general practitioners and practice nurses. This cohort has been followed from baseline for a median of 2.12 years. During that time 2655 people died or were hospitalised with a CVD event. Analyses showed that the original Framingham risk score was reasonably well calibrated overall but underestimated risk in high risk ethnic groups. Discrimination was only modest (AUC 0.701). An adjusted Framingham score, recommended by the New Zealand Guideline Group (NZGG) overestimated 5-year event rates by around 4-7%, in effect lowering the threshold for drug therapy to about 10% 5-year predicted CVD risk. The NZGG adjusted score (AUC 0.676) was less discriminating than the Framingham score and over-adjusted for high risk ethnic groups. For the cohort aged 30-74 years, the NZGG-recommended CVD risk management strategy identified almost half of the population as eligible for lifestyle management +/- drug therapy and this group generated 82% of all CVD events. In contrast the original Framingham score classified less than one-third of the cohort as eligible for individualised management and this group generated 71% of the events that occurred during follow-up. Implications This research project has demonstrated that a CDSS tool can be successfully implemented on a large scale in New Zealand general practice. It has assisted practitioners to improve the assessment and management of CVD at the time of patient consultation. Simultaneously, PREDICT has cost-effectively generated one of the largest cohorts of Māori and non-Māori ever assembled in New Zealand. As the cohort grows, new CVD risk prediction scores will be able to be developed for many New Zealand sub-populations. It will also provide clinicians and policy makers with the information needed to determine the trade-offs between the resources required to manage increasing proportions of the populations and the likely impact of management on preventing CVD events.

Dissertação de mestrado integrado em Engenharia e Gestão de Sistemas de Informação
A Medicina Intensiva (MI) é uma das áreas mais críticas da Medicina. A sua característica multidisciplinar torna-a muito abrangente, reunindo todo o tipo de profissionais de saúde, bem como um local com equipamentos e condições especiais, denominadas Unidades de Cuidados Intensivos (UCI). Tendo em conta o seu ambiente crítico torna-se evidente a necessidade de prever admissões às UCI, pois, para além de constituírem custos adicionais para as instituições e ocuparem recursos desnecessariamente, admissões não planeadas são arriscadas para os doentes que se encontram debilitados. Ao longo dos anos os Sistemas de Informação (SI) têm acompanhando o desenvolvimento da Medicina, tornando-se instrumentos imprescindíveis para o tratamento de doentes, sobretudo através dos Sistemas de Apoio à Decisão (SAD) que apresentam as informações pertinentes sobre os doentes, sem necessidade análise manual de dados. Deste modo, a utilização de SAD na Medicina é crucial, principalmente na MI, em que as decisões têm, muito frequentemente, de ser tomadas com celeridade sempre no melhor interesse do doente. Um SAD pode ser constituído por diferentes técnicas, como é o caso do Data Mining (DM). A presente dissertação envolve descoberta de conhecimento em bases de dados extraídas a partir do sistema de apoio à decisão INTCare, localizado no Centro Hospitalar do Porto (CHP). Foi utilizado um conjunto de técnicas de DM, nomeadamente Clustering e Classificação, tendo por base diferentes algoritmos e métricas de avaliação. Assim foram descobertos padrões naturais nos dados, nomeadamente através da formação de dois grupos de características (Clusters) dos doentes internados em UCI e identificando os atributos mais críticos nestes Clusters. Além disso, foram obtidas previsões com cerca de 97% de capacidade de acertar nos doentes internados (sensibilidade) e que, apesar de criar demasiados Falsos Positivos (63% de especificidade), permitiu obter modelos que permitam que os médicos possam agir de forma proactiva e preventiva, tendo sido esta uma das principais motivações desta dissertação. A presente dissertação serviu para aumentar o número de estudos que aplicam técnicas de DM em MI, particularmente para realização de previsão de internamentos em UCI. Deste modo, contribui-se com conhecimento para a comunidade científica não só de DM, mas também para a Medicina, de modo a potenciar o processo de tomada de decisão médica e na procura pela melhoria dos serviços prestados aos doentes.
Intensive Medicine is one of the most critical areas of medicine. Its multidisciplinary feature makes it a very wide area that gathers all kinds of health professionals as well as a place with special equipment and conditions known as Intensive Care Unit. Having in account its critical environment it becomes evident the need to forecast Intensive Care Unit admissions because, besides being additional costs for institutions and occupy resources unnecessarily, unplanned admissions are risky for patients who are debilitated. Over the years, Information Systems are accompanying the development of medicine and have become essentials instruments for the treatment of patients especially using Clinical Systems Decision Support that have relevant information about patients without the need to manually analyse clinical data. Therefore, the use of DSS is crucial in medicine, particularly in the IM in which decisions must very often be taken speedily always in the best interest of the patient. This Decision Support Systems may be constituted by different techniques such as Data Mining (DM). This dissertation involves knowledge discovery in databases extracted from the Clinical Decision Support System being used in Centro Hospital do Porto (CHP) and named INTCare System. It was used a set of DM and rating techniques including clustering and classification which are based on different algorithms and evaluation metrics. Thereby, natural patterns were discovered in the data particularly through the formation of two groups of characteristics (clusters) of patients admitted to Intensive Care Unit and through the identification the most critical attributes in these clusters. Moreover, it was obtained predictions with approximately 97% of ability to get properly forecast admissions to Intensive Care Unit (Sensitivity) and despite creating too many false positives (63% specificity) it also created models that allow doctors to act proactively and preventively which is one of the main motivations of this dissertation. This dissertation served to increase the number of studies that apply DM techniques in Intensive Medicine particularly for performing predictions of admissions to Intensive Care Units. Thus, knowledge was created for the scientific community not only of DM, but also of medicine in order to promote the process of clinical decision-making and to improve services rendered to patients.

Dissertação de mestrado integrado em Engenharia e Gestão de Sistemas de Informação
Os modelos de apoio à decisão na medicina intensiva são desenvolvidos para apoiar as equipas médicas na tomada de decisão sobre os tratamentos a aplicar a um doente. Existem inúmeros sistemas de apoio à decisão (SAD) que foram desenvolvidos nas últimas décadas para os mais variados ambientes. Em muitos desses SADs, o Machine Learning é utilizado para dar resposta a um problema específico. No entanto, a otimização desses sistemas é particularmente difícil de aplicar devido à dinâmica, complexidade e naturezas multidisciplinares. Com isso, hoje em dia existe uma constante investigação e desenvolvimento de novos algoritmos capazes de extrair conhecimento tratado de grandes volumes de dados, obtendo assim melhores resultados preditivos do que os atuais algoritmos. Existe e emerge um vasto grupo de técnicas e modelos que melhor se adaptam à natureza e complexidade do problema. É nesse propósito que se insere este trabalho. Esta dissertação teve como principal objetivo identificar essas técnicas de otimização, avaliar, comparar e classificar aquelas que melhor podem responder às particularidades da Medicina Intensiva. Como exemplo foram analisados modelos Evolutionary Crisp Rule Learning, Lazy Learning, Evolutionary Fuzzy Rule Learning, Prototype Generation, Fuzzy Instance Based Learning, Decision Trees, Crisp Rule Learning, Neural Networks e Evolutionary Prototype Selection. De seguida foram efetuados alguns desenvolvimentos / testes de modo a aplicar a melhor técnica a um problema de cuidados intensivos, onde a técnicas Decision Trees Genetic Algorithm, Supervised Classifier System e KNNAdaptive obtiveram a melhor taxa de acuidade, mostrando assim a sua exequibilidade e capacidade de atuar em um ambiente real.
The decision support models in intensive care are developed to support medical staff in decision making about treatments to be applied to a patient. There are numerous systems for decision support (DSS) that have been developed in recent decades for a variety of environments. In many of these DSS, the Machine Learning is used to address a specific problem. However, the optimization of these systems is particularly difficult to apply due to the dynamic, complex and multidisciplinary nature. Thus, there is a constant research and development of new algorithms capable of extracting knowledge treated large volumes of data today, able to obtain better predictive results than current algorithms. In fact, emerges a large group of techniques and models that are best suited to the nature and complexity of the problem. This work is incorporated in this context. This dissertation aims to identify these optimization techniques, evaluate, compare and classify them in order to identify what are the best respond to the particularities of Critical Care Medicine. As an example several models were analyzed: Evolutionary Fuzzy Rule Learning, Lazy Learning, Evolutionary Crisp Rule Learning, Prototype Generation, Fuzzy Instance Based Learning, Decision Trees, Crisp Rule Learning, Neural Networks and Evolutionary Prototype Selection. Afterwards some developments / tests were made in order to apply the best technique to a problem of intensive care, where the Decision Trees Genetic Algorithm, Supervised Classifier System and KNNAdaptive obtained the most accurate rate, thus showing their feasibility and ability to work in a real environment.

This thesis study evaluates access to care at an internal medicine unit with solely semi-private rooms at Baystate Medical Center (BMC). Patients are divided into two types: Type I patient consumes one bed; Type II patient occupies two beds or an entire semi-private room as a private space for clinical reasons, resulting in one empty but unavailable (blocked) bed per Type II patient. Because little data is available on blocked beds and Type II patients, unit-level hospital bed planning studies that consider blocked beds have been lacking. This thesis study bridges that gap by building a single-stream and a two-stream discrete micro-simulation model in Excel VBA to describe unit-level bed queue dynamics at hourly granularity in the next 48-hour time horizon, using historical arrival rates and census-dependent discharge rates, supplemented with qualitative results on complexity of patient-level discharge prediction. Results showed that while we increase additional semiprivate beds, there was notable difference between the traditional single-stream model and the two-stream model concerning improvement in bed queue size. Possible directions for future research include patient-level discharge prediction considering both clinical and nonclinical milestones, and strategic redesign of hospital unit(s) considering overflows and internal transfers.

Dissertação de mestrado integrado em Engenharia e Gestão de Sistemas de Informação
Em Medicina Intensiva, a apresentação de informação médica nas Unidades de Cuidados Intensivos (UCI) é feita de diversas formas (gráficos, tabelas, texto, …), pois depende do tipo de análises realizadas, dos dados recolhidos em tempo real pelos sistemas de monitorização, entre outros. A forma como é apresentada a informação pode dificultar a leitura da condição clínica dos doentes por parte dos profissionais de saúde, principalmente quando há a necessidade de um cruzamento entre vários tipos de dados clínicos/fontes de informação. A evolução das tecnologias para novos padrões como a ubiquidade e o pervasive torna possível a recolha e o armazenamento de vários tipos de informação, possibilitando um acesso em temporeal sem restrições de espaço e tempo. A representação de timelines em papel transformou-se em algo desatualizado e por vezes inutilizável devido às diversas vantagens da representação em formato digital. O uso de Sistemas de Apoio à Decisão Clínica (SADC) em UCI não é uma novidade, sendo que a sua principal função é facilitar o processo de tomada de decisão dos profissionais de saúde. No entanto, a associação de timelines a SADC, com o intuito de melhorar a forma como a informação é apresentada, é uma abordagem inovadora, especialmente nas UCI. Este trabalho procura explorar uma nova forma de apresentar a informação relativa aos doentes, tendo por base o espaço temporal em que os eventos ocorrem. Através do desenvolvimento de uma Pervasive Patient Timeline interativa, os profissionais de saúde terão acesso a um ambiente, em tempo real, onde podem consultar o historial clínico dos doentes, desde a sua admissão na unidade de cuidados intensivos até ao momento da alta. Torna-se assim possível visualizar os dados relativos a sinais vitais, análises clínicas, entre outros. A incorporação de modelos de Data Mining (DM) produzidos pelo sistema INTCare é também uma realidade possível, tendo neste âmbito sido induzidos modelos de DM para a previsão da toma de vasopressores, que foram incorporados na Pervasive Patient Timeline. Deste modo os profissionais de saúde passam assim a ter uma nova plataforma capaz de os ajudar a tomarem decisões de uma forma mais precisa.
In Intensive Care Medicine, the presentation of medical information in the Intensive Care Units (ICU) is done in many shapes (graphics, tables, text,…). It depends on the type of exams executed, the data collected in real time by monitoring systems, among others. The way in which information is presented can make it difficult for health professionals to read the clinical condition of patients. When there is the need to cross between several types of clinical data/information sources the situation is even worse. The evolution of technologies for emerging standards such as ubiquity and pervasive makes it possible to gather and storage various types of information, thus making it available in real time and anywhere. Also with the advancement of technologies, the representation of timelines on paper turned into something outdated and sometimes unusable due to the many advantages of representation in digital format. The use of Clinical Decision Support Systems (CDSS) is not a novelty, and its main function is to facilitate the decision-making process, through predictive models, continuous information monitoring, among others. However, the association of timelines to CDSS, in order to improve the way information is presented, is an innovative approach, especially in the ICU. This work seeks to explore a new way of presenting information about patients, based on the time frame in which events occur. By developing an interactive Pervasive Patient Timeline, health professionals will have access to an environment in real time, where they can consult the medical history of patients. The medical history will be available from the moment in which patients are admitted in the ICU until their discharge, allowing health professionals to analyze data regarding vital signs, medication, exams, among others. The incorporation of Data Mining (DM) models produced by the INTCare system is also a reality, and in this context, DM models were induced for predicting the intake of vasopressors, which were incorporated in Pervasive Patient Timeline. Thus health professionals will have a new platform that can help them to make decisions in a more accurate manner.