Presepsin (PSEP) is released during infectious diseases and can be detected in the blood. PSEP has shown promising results as sepsis marker. We examined the diagnostic and prognostic validity of PSEP in patients suspicious of sepsis on admission in the emergency department (ED). Methods One hundred twenty three patients with signs of SIRS and/or sepsis and 123 healthy individuals were enrolled. PSEP was determined on admission, after 8, 24 and 72 h. Results Mean PSEP concentrations of the control group and the patient group were 130 and 1945 pg/ml. PSEP differed between SIRS, sepsis, severe sepsis and septic shock and showed strong association with 30-day mortality ranging from 10.3% in the 1st to 32.1% in the 4th quartile. The ROC curve analyses revealed an AUC value of 0.743. Combined assessment of PSEP and MEDS score increased the AUC up to 0.878 demonstrating the close relationship with outcome. Based on the PSEP values in the different severity degrees, decision thresholds for risk stratification were established. The course of PSEP during the first 72 h was associated with effectiveness of treatment and outcome. Conclusions PSEP allowed outcome prediction already on admission to a similar degree as the clinical scores MEDS and APACHE II. Combination of PSEP with MEDS score improved the discriminatory power for outcome prediction. / Our study has been supported by Mitsubishi Chemical Europe through providing the PSEP reagents free of charge. Dr. Carpio has received speaker honoraria from Mitsubishi Chemical Europe. DIAneering – Diagnostics Engineering & Research consulted to Axis Shield Diagnostics, Mitsubishi Chemical Europe, Radiometer, Roche Diagnostics, Shanghai Kehua Bio-engineering. No potential conflict of interest to this paper was reported / Peer review

Because tobacco smoking causes 430,000 U.S. deaths annually, wide-reaching smoking bans are needed. Bans reduce cigarette consumption, encourage cessation, protect nonsmokers from second-hand smoke, and promote an attitude that smoking is undesirable. Therefore, bans may prevent future generations from suffering many smoking-related health problems. The federal government has not implemented widereaching smoking bans so it falls on individual states, counties, or communities to devise appropriate smoking policy. To date, smoking policy has been determined by legislators, who may have conflicts that prevent them from acting in the publics best interest. However, this method of implementing smoking policy may be changing. In 2005, Washington residents voted by ballot initiative to strengthen existing state smoking regulations. In 2006, Arizona, Nevada, and Ohio residents voted by ballot initiatives to implement strict statewide smoking bans. This research presents a way to predict how residents of other states might vote if given the opportunity. Two research hypotheses are tested and accepted: a positive relationship between favorable votes and urbanness, and a preference favoring smoking bans where smoking regulations already exist. Finally, a projection is made that a smoking ban vote in Kentucky would yield favorable results, and a map showing projected county votes is provided.

The main idea of this thesis is to present the novel concept of localized feature selection (LFS) for data classification and its application for coma outcome prediction. Typical feature selection methods choose an optimal global feature subset that is applied over all regions of the sample space. In contrast, in this study we propose a novel localized feature selection approach whereby each region of the sample space is associated with its own distinct optimized feature set, which may vary both in membership and size across the sample space. This allows the feature set to optimally adapt to local variations in the sample space. An associated localized classification method is also proposed. The proposed LFS method selects a feature subset such that, within a localized region, within-class and between-class distances are respectively minimized and maximized. We first determine the localized region using an iterative procedure based on the distances in the original feature space. This results in a linear programming optimization problem. Then, the second method is formulated as a non-linear joint convex/increasing quasi-convex optimization problem where a logistic function is applied to focus the optimization process on the localized region within the unknown co-ordinate system. This results in a more accurate classification performance at the expense of some sacrifice in computational time. Experimental results on synthetic and real-world data sets demonstrate the effectiveness of the proposed localized approach. Using the LFS idea, we propose a practical machine learning approach for automatic and continuous assessment of event related potentials for detecting the presence of the mismatch negativity component, whose existence has a high correlation with coma awakening. This process enables us to determine prognosis of a coma patient. Experimental results on normal and comatose subjects demonstrate the effectiveness of the proposed method. / Dissertation / Doctor of Philosophy (PhD) / This study proposes a novel form of pattern classification method, which is formulated in a way so that it is easily executable on a computer. Two different versions of the method are developed. These are the LFS (localized feature selection) and lLFS (logistic LFS) methods. Both versions are appropriate for analysis of data with complex distributions, such as datasets that occur in biological signal processing problems. We have shown that the performance of the proposed methods is significantly improved over that of previous methods, on the datasets that were considered in this thesis. The proposed method is applied to the specific problem of determining the prognosis of a coma patient. The viability of the formulation and the effectiveness of the proposed algorithm are demonstrated on several synthetic and real world datasets, including comatose subjects.

Contexte : il est possible qu’une victime subisse de multiples conséquences d’accident de la route,conséquences pouvant retentir durablement sur sa vie. Cependant, peu d’études permettent de connaitre leprofil du blessé grave ainsi que les facteurs prédictifs de son devenir. De plus, il existe peu d'outilsprédictifs servant à prédire les conséquences post-accidentelles. L’objectif de la thèse est de caractériserces conséquences, de chercher les éléments pronostiques de gravité des conséquences un an aprèsl’accident et de donner une évaluation, à partir de données réelles, de la qualité de prédiction del’indicateur de déficience à un an appelé IIS (Injury Impairment Score - un indice de déficience - défini apriori à partir des lésions- et utilisé fréquemment).Méthodes : la thèse est réalisée dans le cadre de la cohorte ESPARR (Étude et Suivi d’une Populationd’Accidentés de la Route dans le Rhône), qui s’appuie sur les données du Registre des accidents de lacirculation du Rhône, et qui inclut 1372 sujets blessés dans des accidents de la route dont 1168 sujets âgésde 16 ans et plus. Parmi ces sujets, 886 adultes ont répondu à un questionnaire de suivi à un an, 616 sujetsont des données complètes et sont classés dans des groupes homogènes en fonction de leur devenir à unan par l'analyse des correspondances multiples et la méthode de classification hiérarchique. L’analyse desfacteurs prédictifs de leur appartenance à un de ces groupes de victimes, mesurés à la date de l'accident, aété effectuée à l’aide de modèles de régressions logistiques multinomiales pondérés. L'évaluation de l'IISsur les données réelles est réalisée en regardant la cohérence entre l'IIS et les différents facteurs mesurés àun an.Résultats : cinq groupes homogènes au niveau des conséquences de l’accident à un an ont été identifiés :le groupe-1 contient 206 sujets, dont une majorité est considérée en bonne récupération ; le groupe-2concerne les sujets ayant uniquement des conséquences physiques ; les groupes 3, 4 et 5 concernent lessujets ayant des conséquences multiples. À part les conséquences physiques en lien avec les sujets dansces groupes, certains plus en lien avec des répercussions sur la vie social (groupe-3), d’autres en lien avecdes difficultés sociales ou environnementales (groupe-4). Le groupe-5 comprend tous les sujets quisouffrent de syndrome post-commotionnel de la population d’étude. Après avoir ajusté sur plusieursvariables recueillies lors de l'accident, notre étude montre que, en plus des facteurs déjà évoqués dans lalittérature (âge, gravité…), le niveau de fragilité socioéconomique et le fait d'avoir un proche blessé dansl'accident sont également des facteurs prédisant le devenir des victimes d’un accident. En ce qui concernel'évaluation de l'IIS sur les données réelles, nous trouvons que le niveau des conséquences prédites par l’IIS ne correspond pas parfaitement à celui observé en réalité à un an quels que soient les facteurs mesurés.Conclusion : un an après l’accident, de nombreuses victimes d'accident de la route, même parmi cellessouffrant de lésions légères, continuent de présenter de multiples problèmes tant sur leur santé physiqueque mentale, sur le plan social ainsi que sur leur environnement. Dans une perspective de réadaptation à lavie quotidienne, ces résultats peuvent être utiles à l’amélioration de la prise en charge des accidentés de laroute. / Background: it is possible that victims can suffer from multiple problems after an accident, and this canbe seen in the people with the most serious consequences. However, few studies allow us to know theprofile and prognostic factors of severity of consequences after the accident in this population of victims.Moreover, there are few tools to predict 1-year post-traumatic sequelae in road crash victims.The thesis aims to determine subgroups of victims with similar outcomes 1 year after the crash andpredictive factors for attribution to these subgroups and validate sequelae prediction by the InjuryImpairment Score (IIS), in comparison with the one year outcomes.Methods: the thesis is a part of the broader ESPARR study based on the Rhône Registry of Road TrafficCasualties. The ESPARR cohort comprised 1,372 subjects, including 1,168 aged 16 years. Among 886adult subjects who responded to a follow-up questionnaire one year later, the main analysis was carriedout on 616 participants, who completed a self-report questionnaire on health, social, emotional andfinancial status 1 year after a crash. The multiple correspondence analysis and hierarchical clusteringmethod was implemented to produce homogeneous road-crash victim subgroups according to differencesin outcome. Baseline (time of accident) predictive factors for subgroup attribution were analysed onweighted multinomial logistic regression models. We used outcomes data at 1-year follow-up of roadinjury to validate the ability of IIS to predict sequelae.Results: five different victim groups were identified in terms of consequences one year after the crash:one group (206 subjects, 33.4%) presented few problems, one group with essentially physical sequelae,one group with essentially physical and social problems, and two groups presented many problems (oneincluded more victims with psychological problem and less environment problem). As well as the knownprognostic factors of age, initial injury severity and lesion type, socioeconomic fragility and the fact of arelative being involved in the accident emerged as being predictive of poor outcome one year later. IIS, inthis injured population, failed to predict sequelae one year later as measured by real data.Conclusion: one year after a road accident, victims may still experience multiple problems in terms notonly of physical health but also of mental health, social life and environment. Poor outcome may bepredicted both from accident-related factors and from victims' socioeconomic fragility. These findings areuseful in guiding prevention in terms not only of recovery of health status but also of recovery of sociallife in the best possible environment.

Cervical cancer remains the third most commonly diagnosed gynecological malignancy in the United States and throughout the world despite being potentially preventable. Patients diagnosed with cervical cancer may develop local recurrence in the cervix and surrounding structures (vaginal apex, parametrial, or paracervical), regional recurrence in pelvic lymph nodes, distant metastasis, or a combination of all. The management of such treatment outcomes has not been subject to rigorous investigation. Therefore, there is a need for studies and clinical trials that focus on decision making to support the choice of the best treatment modality that leads to the minimal number of adverse treatment outcomes. Medical imaging plays a vital role in the initial diagnosis, staging, and guiding treatment decisions for cancer patients. Positron Emission Tomography-Computed Tomography (PET/CT) hybrid scanner has proven to be a primary functional imaging modality in the oncology clinic. A typical oncological application of PET/CT aims to examine the whole body for high tracer uptake as a sign of tumorous lesions or metastasis using 18F-Fluoro-2-deoxy-D-glucose (18F-FDG). This radiopharmaceutical has been proven to be useful for the quantitative determination of regional glucose metabolism localized in the brain, heart, bladder, and, fortunately, in tumors. Currently, 18F-FDG measured on PET is the prominent radiotracer in cancer staging and follow-up imaging. In the –omics1 era, mining data to derive inherent information about a system has influenced the medical field, especially oncological imaging. The process of radiomics involves high throughput analysis of medical images to extract a large number of quantified features that are presented as a decision supporting tool for clinicians in terms of various clinical tasks such as staging, prediction, and prognosis. In recent studies, the focus of radiomics has exceeded the whole-tumor analysis to include the quantification of habitats, sub-regions within the tumor volume defined based on specific criteria, with the intent to investigate the diversity extent of the intratumor heterogeneity as robust descriptors and predictors of clinicopathological factors. The presented work is a retrospective analysis of a cohort consisting of pretreatment Positron Emission Tomography and Computed Tomography (PET/CT) hybrid scans of cervical cancer patients consecutively treated with radiochemotherapy. We extracted radiomic features from the primary cervical tumor volumes, and voxel intensity-based features from tumor habitats to analyze the tumors’ heterogeneity based on 18Flourodeoxyglocuse (18F-FDG) uptake of PET, and Hounsfield Units (HU) of CT to obtain useful tumor information, which might be associated with treatment outcomes. To our knowledge, a limited number of studies have focused on investigating the potential role of radiomic features on cervical cancer PET/CT images. Briefly, the workflow of this study consisted of investigating parameters that might affect radiomic features predictive performance by evaluating the reproducibility of radiomic features extracted from 18F-FDG PET images for segmentation methods, gray levels discretization, and PET reconstruction algorithms. Afterward, we used these features to predict cervical treatment outcomes after radiochemotherapy. Due to the use of human data, this research study acquired the approval of the institutional review board (IRB) at the University of South Florida.

Pain from the musculoskeletal system is very common in the modern society. Chronic musculoskeletal pain syndromes causes not only individual suffering but also dysfunctions of movements and postural control, as large costs for the society. In spite of significant efforts, there is a shortage of knowledge on effective prevention, diagnoses and rehabilitation of different chronic musculoskeletal pain syndromes. The general aims of this thesis was to investigate the predictive value of physical, sociodemographic, and psychosocial-behavioural variables for pain reduction after multimodal rehabilitation in patients with chronic low back or neck pain, and to develop and evaluate tests for objective and quantitative evaluation of characteristic sensorimotor disturbances in chronic neck pain. Logistic regression models revealed that unchanged pain intensity could be predicted with good precision while reduced pain intensity after rehabilitation was poorly predicted by the baseline variables. Altered pain intensity in chronic low back pain was predicted by high pain intensity, low levels of pain severity and high affective distress, while reduced pain intensity for patients with chronic neck pain were predicted by high endurance, low age, high pain intensity, low need of being social along with optimistic attitudes on how the pain will interfere with daily life, and few vegetative symptoms. One of the conclusions was that objective measures of specific sensorimotor disturbances should improve the precision by which treatment-induced effects can be assessed and predicted. A study was designed to objectively and quantitatively evaluate a large numbers of different sensorimotor characteristics in a small group of patients with chronic neck pain of different aetiology (whiplash-related and insidious). Kinematic data was recorded during different motor tasks, involving cervical rotations, arm movements and standing. In comparison to a group of asymptomatic control subjects, patients with chronic neck pain was characterised by slower movements, poor balance, reduced cervical stability during perturbations, altered smoothness of movement (jerk index), and reduced movement precision (variable error and variability in range of motion). The sensorimotor variables velocity of arm movements and cervical stability, could correctly classified nearly 90% of the subjects as having chronic neck pain or being asymptomatic. There was a large diversity of sensorimotor disturbances among the individual patients. This was confirmed in a regression model that failed to separate the groups insidious neck pain (sensitivity 44%) and WAD (sensitivity 67%). By investigating associations between the different sensorimotor variables, close relations was found between the repositioning acuity and variability in range of motion, and between standing balance and cervical stability/ standing balance during perturbation. These two groups of variables were only weakly related to each other and to smoothness of movement and movement velocity. The results indicate that chronic neck pain is characterised by specific sensorimotor deficits, and that there are common pathophysiological mechanisms in chronic neck pain of different aetiology. However, the lack of associations between several sensorimotor disturbances indicates that different mechanisms are involved. The thesis indicates that objective sensorimotor tests should be used to improve the quality of functional assessments in chronic neck pain. Methods that objectively and quantitatively measure e.g. movement precision, balance and cervical stability are also needed in order to evaluate current treatment methods and to develop new rehabilitation programs for specific sensorimotor deficits.

Motivation: Disease progression in cancer can vary substantially between patients. Yet, patients often receive the same treatment. Recently, there has been much work on predicting disease progression and patient outcome variables from gene expression in order to personalize treatment options. A widely used approach is high-throughput experiments that aim to explore predictive signature genes which would provide identification of clinical outcome of diseases. Microarray data analysis helps to reveal underlying biological mechanisms of tumor progression, metastasis, and drug-resistance in cancer studies. Despite first diagnostic kits in the market, there are open problems such as the choice of random gene signatures or noisy expression data. The experimental or computational noise in data and limited tissue samples collected from patients might furthermore reduce the predictive power and biological interpretability of such signature genes. Nevertheless, signature genes predicted by different studies generally represent poor similarity; even for the same type of cancer. Integration of network information with gene expression data could provide more efficient signatures for outcome prediction in cancer studies. One approach to deal with these problems employs gene-gene relationships and ranks genes using the random surfer model of Google's PageRank algorithm. Unfortunately, the majority of published network-based approaches solely tested their methods on a small amount of datasets, questioning the general applicability of network-based methods for outcome prediction. Methods: In this thesis, I provide a comprehensive and systematically evaluation of a network-based outcome prediction approach -- NetRank - a PageRank derivative -- applied on several types of gene expression cancer data and four different types of networks. The algorithm identifies a signature gene set for a specific cancer type by incorporating gene network information with given expression data. To assess the performance of NetRank, I created a benchmark dataset collection comprising 25 cancer outcome prediction datasets from literature and one in-house dataset. Results: NetRank performs significantly better than classical methods such as foldchange or t-test as it improves the prediction performance in average for 7%. Besides, we are approaching the accuracy level of the authors' signatures by applying a relatively unbiased but fully automated process for biomarker discovery. Despite an order of magnitude difference in network size, a regulatory, a protein-protein interaction and two predicted networks perform equally well. Signatures as published by the authors and the signatures generated with classical methods do not overlap -- not even for the same cancer type -- whereas the network-based signatures strongly overlap. I analyze and discuss these overlapping genes in terms of the Hallmarks of cancer and in particular single out six transcription factors and seven proteins and discuss their specific role in cancer progression. Furthermore several tests are conducted for the identification of a Universal Cancer Signature. No Universal Cancer Signature could be identified so far, but a cancer-specific combination of general master regulators with specific cancer genes could be discovered that achieves the best results for all cancer types. As NetRank offers a great value for cancer outcome prediction, first steps for a secure usage of NetRank in a public cloud are described. Conclusion: Experimental evaluation of network-based methods on a gene expression benchmark dataset suggests that these methods are especially suited for outcome prediction as they overcome the problems of random gene signatures and noisy expression data. Through the combination of network information with gene expression data, network-based methods identify highly similar signatures over all cancer types, in contrast to classical methods that fail to identify highly common gene sets across the same cancer types. In general allows the integration of additional information in gene expression analysis the identification of more reliable, accurate and reproducible biomarkers and provides a deeper understanding of processes occurring in cancer development and progression.

<p>The present dissertation comprises four empirical studies within the area of behavioral treatment of panic disorder with agoraphobia. The focus is on studying issues pertaining to outcome, treatment process and determinants of change. The first study is a randomized controlled treatment study of 73 patients undergoing 16 sessions of either exposure in vivo (E), or cognitive behavior therapy (CBT). Both treatments showed clear improvements at post-treatment that were well maintained at 1-year follow up, and there were no significant differences between the treatments.</p><p>The second study concerned prediction of outcome in the same sample. From a variety of pre-treatment characteristics severity of avoidance was the one most related to outcome. Most predictors were found unrelated. Two approaches of prediction were also compared: treating outcome as a categorical vs. continuous variable. The different approaches yielded a somewhat dissimilar picture of the impact of pre-treatment severity of avoidance. The third study examined different aspects of the therapeutic relationship, and their relation to outcome. Clients’ perceptions of therapists and their ratings of the working alliance were generally not related to outcome at any point. On the other hand, therapists’ perceptions of patients as showing goal-direction and active participation were related to outcome from early on in therapy. The fourth study examined different aspects of change. It was found that change in indices of the frequency of panic attacks was not closely related to change in agoraphobic avoidance at post-treatment. Change in avoidance was also more related to other aspects of outcome. At one-year follow-up, a more unitary picture, regarding the different aspects of change was observed.</p>

The present dissertation comprises four empirical studies within the area of behavioral treatment of panic disorder with agoraphobia. The focus is on studying issues pertaining to outcome, treatment process and determinants of change. The first study is a randomized controlled treatment study of 73 patients undergoing 16 sessions of either exposure in vivo (E), or cognitive behavior therapy (CBT). Both treatments showed clear improvements at post-treatment that were well maintained at 1-year follow up, and there were no significant differences between the treatments. The second study concerned prediction of outcome in the same sample. From a variety of pre-treatment characteristics severity of avoidance was the one most related to outcome. Most predictors were found unrelated. Two approaches of prediction were also compared: treating outcome as a categorical vs. continuous variable. The different approaches yielded a somewhat dissimilar picture of the impact of pre-treatment severity of avoidance. The third study examined different aspects of the therapeutic relationship, and their relation to outcome. Clients’ perceptions of therapists and their ratings of the working alliance were generally not related to outcome at any point. On the other hand, therapists’ perceptions of patients as showing goal-direction and active participation were related to outcome from early on in therapy. The fourth study examined different aspects of change. It was found that change in indices of the frequency of panic attacks was not closely related to change in agoraphobic avoidance at post-treatment. Change in avoidance was also more related to other aspects of outcome. At one-year follow-up, a more unitary picture, regarding the different aspects of change was observed.

Advances in high-throughput genomic and proteomic technology have led to a growing interest in cancer biomarkers. These biomarkers can potentially improve the accuracy of cancer subtype prediction and subsequently, the success of therapy. However, identification of statistically and biologically relevant biomarkers from high-throughput data can be unreliable due to the nature of the data--e.g., high technical variability, small sample size, and high dimension size. Due to the lack of available training samples, data-driven machine learning methods are often insufficient without the support of knowledge-based algorithms. We research and investigate the benefits of using knowledge-based algorithms to solve clinical prediction problems. Because we are interested in identifying biomarkers that are also feasible in clinical prediction models, we focus on two analytical components: feature selection and predictive model selection. In addition to data variance, we must also consider the variance of analytical methods. There are many existing feature selection algorithms, each of which may produce different results. Moreover, it is not trivial to identify model parameters that maximize the sensitivity and specificity of clinical prediction. Thus, we introduce a method that uses independently validated biological knowledge to reduce the space of relevant feature selection algorithms and to improve the reliability of clinical predictors. Finally, we implement several functions of this knowledge-based method as a web-based, user-friendly, and standards-compatible software application.