The selection of the best response variables in a clinical trial is often not straightforward; the primary endpoint of a trial should be clinically relevant, directly related to the primary objective of the trial, and with favorable efficiency to detect the treatment benefit with a reasonable sample size and duration of the trial. With the recent success in the management of heart failure, the mortality rate has dropped significantly compared to two decades ago, and patients with heart failure have high rates of hospitalization and morbid complications along with multiple symptoms and severe limitations in daily activities. Although mortality still remains important as a measure of the clinically relevant benefit and the safety of the intervention, with the low event rate of mortality, it requires large and longer clinical trials to detect treatment benefit of new intervention using mortality as the sole primary endpoint. Thus most heart failure trials use the combined endpoint of death and a second efficacy outcome, such as hospitalizations. This is often analyzed with time-to-first-event survival analysis which ignores possible subsequent hospitalization events and treating the death and first hospitalization equally in the importance and hierarchy of clinical relevance. Accounting for the recurrent events or subsequent death after the hospitalization(s) provides more detailed information on the disease-control process and treatment benefit. In this dissertation we propose a hierarchical endpoint with death in the higher priority and number of hospitalization events in the lower priority as primary endpoint to assess experimental treatment benefit versus a control using a non-parametric generalized Gehan-Wilcoxon test. In addition to the hierarchical endpoint, we also evaluated assessment of experimental treatment benefit on recurrent events with a multi-state model using extended stratified Cox model, considering the multi-states in which patients might transition during the study. We compared the false positive rate and power of the above mentioned methods with the composite endpoint approach and recurrent event endpoint approach analyzed using Andersen-Gill, WLW, and PWP models in simulation studies. Finally we applied all evaluated procedures to the Digitalis Investigation Group (DIG) trial.

Gestational diabetes mellitus (GDM) is the most common complication of pregnancy that affects 1-14% of all pregnancies. If not properly managed, GDM can be a devastating disease, leading to birth complications such as shoulder dystocia and neonatal hypoglycemia. GDM has many long-term implications as well, such as increased risk of obesity and type 2 diabetes mellitus (T2DM) in both the mother and the offspring. Additionally, women with a history of GDM are at increased risk of recurrent GDM in a subsequent pregnancy and multiple episodes of GDM further increases a woman’s risk for these short and long-term consequences. For this reason, a diagnosis of GDM provides an opportunity to target GDM and T2DM risk factors to prevent recurrence of GDM and halt the diabetes disease course. Research has shown that diet and physical activity interventions provided after a pregnancy complicated by GDM can delay or prevent the onset of T2DM yet literature on prevention of recurrent GDM is lacking. This thesis will propose a new intervention applied to the inter-pregnancy interval (IPI), designed to reduce incidence of recurrent GDM. The study will examine the effect of a diet and physical activity intervention for women with a recent pregnancy complicated by GDM on recurrence in a subsequent pregnancy and weight gain in the IPI. We hypothesize that our intervention will reduce incidence of GDM recurrence compared to the control group and that women in the intervention group will lose more weight compared to women in the control group. The results of this study will provide a background for further study on the prevention of GDM recurrence with the hope that prevention of recurrent GDM will prevent the short and long-term sequela of GDM.

Classical novae (CNe) are interacting binary systems whose outbursts are powered by a thermonuclear runaway in accreted material on the surface of a white dwarf (WD). The secondary stars in such systems fill their Roche lobe and material is transferred onto the WD primary star via an accretion disk. Recurrent novae (RNe) show many similarities to CNe, but have had more than one recorded outburst. They play an important role as one of the suspected progenitor systems of Type Ia supernovae (SNe) which are used as primary distance indicators in cosmology. Thus, it is important to investigate the nature of their central binary systems to determine the relation between the parameters of the central system and outburst type, and finally ascertain the population of novae that might be available to give rise to the progenitors of Type Ia SNe. The investigation looking for characteristics that may distinguish RNe from CNe systems, the selection of initial targets for detailed study, and results of the investigation are presented in this thesis. The proposal that RNe occupy a region separated from CNe in an outburst amplitude versus speed class diagram was adopted. Since the low amplitude results from the existence of an evolved secondary and/or high mass transfer rate in the quiescent system, RNe candidates should accordingly have low amplitude. The 93 novae with observed V amplitudes given in the literature and 43 novae with photographic amplitudes have been combined and plotted on an outburst amplitude versus rate of decline diagram from which 16 target novae suspected to be RNe candidates were selected for photometric and spectroscopic follow-up. Quiescent photometric magnitudes and spectra were obtained using RATCam on LT, FRODOSpec on LT, and RSS on SALT. Spectral type and luminosity class determined from the near-IR colour-magnitude diagrams were compared to those derived from the spectra. Determination of spectral types was accomplished by identifying specific lines and calculating indices from TiO bands, VO bands, and the Na atomic line for giants (finding 4 stars) and sub-giants/giants (3 stars). A spectral library template was used instead of the indices in cases of main-sequence stars (2 stars).Our investigation also confirmed the positions of AR Cir, V794 Oph and EU Sct where there had been some ambiguity previously. Ultimately, we suggest four prime RNe candidates (2 novae with giant secondaries - V3964 Sgr and EU Sct, and 2 novae with sub-giant secondaries - V794 Oph and V368 Aql) which are currently classified as CNe, to look for more than one outburst in archival plates or large sample sky surveys such as SMEI (see below). By introducing the high cadence full-sky space-based observational archive of the Solar Mass Ejection Imager (SMEI) which operated on the Coriolis satellite from 2003-2011, we derived light curves of one Mira (O Cet) as a general example and two novae with known outbursts during 2003-2011 (V2467 Cyg and V1187 Sco). The SMEI light curves potentially reveal more details than those given by ground-based observations. The pre-maximum halt was found in V2467 Cyg as well as oscillations in light curves found earlier than those found in previous studies. The precise date of maximum of each nova was provided. Four bright novae that are potentially RNe candidates (V4074 Sgr, V3964 Sgr, DK Lac and V368 Aql) were searched for second outbursts in the SMEI data, but none were found. Among the nova outbursts detected by SMEI, we found however unprecedented detail in first class data of the Recurrent Nova T Pyx in its 2011 outburst. We investigated the optical light curve of T Pyx during its 2011 outburst through compiling a database of SMEI and American Association of Variable Star Observers (AAVSO) observations. The SMEI light curve, covering t=1.5-49 days post-discovery, was divided into four phases based on the idealised nova optical light curve; the initial rise (1.5-3.3 days), the pre-maximum halt (3.3-13.3 days), the final rise (14.7-27.9 days), and the early decline (27.9 days, until the end of SMEI observations on day 49). The SMEI light curve contains a strongly detected period of 1.44+-0.05 days during the pre-maximum halt phase. These oscillations resemble those found in recent TNR models arising from instabilities in the expanding envelope. No spectral variations that mirror the light curve periodicity were found however. The marked dip at t~22-24 days just before light curve maximum at t=27.9 days may represent the same (shorter duration) phenomenon seen in other novae observed by SMEI and present in some TNR model light curves. The spectra of T Pyx from the 2m the Liverpool Telescope and the Small and Moderate Aperture Research Telescope System (SMARTS) 1.5m telescope were obtained from t=0.8-80.7 and 155.1-249.9 days, covering the major phases of development. The nova was observed very early in its rise where a distinct high velocity ejection phase was evident with derived Vej~4000 km/s initially. A marked drop at t=5.7 days, and then a gradual increase occurred in derived Vej to stabilise at ~1500 km/s at the pre-maximum halt. Here we propose two different stages of mass loss, a short-lived phase occurring immediately after outburst and lasting ~6 days followed by a more steadily evolving and higher mass loss phase. The overall spectral development follows that typical of a Classical Nova and comparison with the photometric behaviour reveals consistencies with the simple evolving pseudo-photosphere model of the nova outburst. Comparing optical spectra to X-ray and radio light curves, weak [Fe X] 6375A emission was marginally detected before the X-ray rise and was clearly present during the brightest phase of X-ray emission. If the onset of the X-ray phase and the start of the final decline in the optical are related to the cessation of significant mass loss, then this occurred at t~90-110 days.

Made available in DSpace on 2014-06-11T19:23:07Z (GMT). No. of bitstreams: 0 Previous issue date: 2007-02-28Bitstream added on 2014-06-13T19:49:52Z : No. of bitstreams: 1 mangerona_lr_me_botfm.pdf: 551674 bytes, checksum: fa33b5ef943c594fe040a66778a352bd (MD5) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Ministério da Saúde / Secretaria do Estado da Saúde de São Paulo / A perda gestacional recorrente idiopática é multifatorial, pois envolve fatores de risco clínicos e biológicos. A trombofilia pode ser definida como uma predisposição para trombose. Anormalidades na hemostasia que estão associadas com trombofilias clínicas incluem defeitos hereditários, tais como os anticoagulantes naturais Antitrombina III, Proteína S e Proteína C ou fatores de coagulação, as mutações do fator V Leiden, gene da protrombina G20210A, metilenotetrahidrofolato redutase MTHFR C677T, e defeitos adquiridos, tal como Síndrome Antifosfolípide e a Hiperhomocisteinemia. O presente trabalho foi realizado com 70 mulheres, sendo que 35 mulheres apresentavam 3 ou mais abortos recorrentes inexplicáveis, e 35 mulheres voluntárias clinicamente normais, para todas as mulheres foram feitas as investigações para os anticoagulantes naturais da coagulação e investigação para as mutações do fator V Leiden, gene da protrombina G20210A e metilenotetrahidrofolato redutase MTHFR C677T e as deficiências adquiridas. Em nosso estudo, encontramos resultados estatísticamente significantes para a síndrome do anticorpo antifosfolípide (trombofilia adquirida). Em nosso estudo observamos um grande número de defeitos trombofílicos adquiridos sendo que alguns estão em associação com a mutação do fator V Leiden e MTHFR C677T, porém mais pesquisas são necessárias para confirmar ou contestar as causas das trombofilias,e avaliar a eficiência e segurança da tromboprofilaxia em mulheres grávidas. / The idiopathic appealing gestational loss is multifactorial because it involves clinical and biological risk factors. Thrombophilia can be defined as a predisposition for thrombosis. Abnormalities in homeostasis that are associated with clinical thrombophilia include hereditary defects, such as natural anticoagulants as Antithrombin III, S Protein and C Protein or coagulation factors, mutations of V Leiden factor, gene of G20210A prothrombin, MTHFR C677T methilene redutase tetrahydropholat, and acquired defects just as Antiphospholipid Syndrome and Hyperhomocisteinemy. The present work was accomplished with 70 women, and 35 women showed 3 or more inexplicable appealing abortions, and 35 women voluntary clinically normal. For all women were made investigations for natural anticoagulants of coagulation and investigations for mutations of V Leiden factor, gene of G20210A prothrombin, MTHFR C677T methilene redutase tetrahydropholat, and acquired defects. In study we found statistically significant results for antiphospholipid antibody syndrome (acquired thrombophilia). In our study it was observed a great number of thrombophilics acquired defects, and some of then are associated to mutation of V Leiden factor and MTHFR C677T, however more researches are necessary to confirm or to answer the causes of the thrombophilia and to evaluate the efficiency and safety of the thromboprophylaxis in pregnant women.

Two classes of recurrent neural network models are compared in this report, simple recurrent networks (SRNs) and sequential cascaded networks (SCNs) which are first- and second-order networks respectively. The comparison is aimed at describing and analysing the behaviour of the networks such that the differences between them become clear. A theoretical analysis, using techniques from dynamic systems theory (DST), shows that the second-order network has more possibilities in terms of dynamical behaviours than the first-order network. It also revealed that the second order network could interpret its context with an input-dependent function in the output nodes. The experiments were based on training with backpropagation (BP) and an evolutionary algorithm (EA) on the AnBn-grammar which requires the ability to count. This analysis revealed some differences between the two training-regimes tested and also between the performance of the two types of networks. The EA was found to be far more reliable than BP in this domain. Another important finding from the experiments was that although the SCN had more possibilities than the SRN in how it could solve the problem, these were not exploited in the domain tested in this project

No description available.

Modern automobiles have been proven vulnerable to hacking by security researchers. By exploiting vulnerabilities in the car's external interfaces, attackers can access a car's controller area network (CAN) bus and cause malicious effects. We seek to detect these attacks on the bus as a last line of defence against automotive cyber attacks. The CAN bus standard defines a low-level message structure, upon which manufacturers layer their own proprietary command protocols; attacks must similarly be tailored for their target. This variability makes intrusion detection methods difficult to apply to the automotive CAN bus. Nevertheless, the bus traffic is generated by machines; thus we hypothesize that it can be characterized with machine learning, and that attacks produce anomalous traffic. Our goals are to show that anomaly detection trained without understanding of the message contents can detect attacks, and to create a framework for understanding how the characteristics of a novel attack can be used to predict its detectability. We developed a model that describes attacks based on their effect on bus traffic, informed by a review of published material on car hacking in combination with analysis of CAN traffic from a 2012 Subaru Impreza. The model specifies three high-level categories of effects: attacks that insert foreign packets, attacks that affect packet timing, and attacks that only modify data within packets. Foreign packet attacks are trivially detectable. For timing-based anomalies, we developed features suitable for one-class classification methods. For packet stream data word anomalies, we adapted recurrent neural networks and multivariate Markov model methods to sequence anomaly detection and compared their performance. We conducted experiments to evaluate our detection methods with special attention to the trade-off between precision and recall, given that a practical system requires a very low false alarm rate. The methods were evaluated by synthesizing anomalies within each attack category, parameterized to adjust their covertness. We generalize from the results to enable prediction of detection rates for new attacks using these methods.

There is generally a lack of studies examining prevalence and phenomenology of bipolar disorder in Africa. In literature, a unipolar manic course of illness in particular is reported to be rare. The purpose of this study was to investigate and describe the course of illness and clinical features in a cross-section of patients diagnosed with bipolar disorder attending public hospitals in Limpopo Province, South Africa and to determine the rate of a unipolar manic course in this sample of patients. This was a descriptive, cross-sectional study of patients presenting with a history of mania between October 2009 and April 2010, to three hospitals in Limpopo Province. A purposeful sample of 103 patients was recruited and interviewed using the Affective Disorders Evaluation. This study confirms that a unipolar manic course is indeed much more common than rates suggested in present day literature with57% of the study sample only ever experiencing manic episodes. The study also confirms the debilitating nature of bipolar disorder with more than two-thirds being unemployed in spite of a quarter of the study subjects having a tertiary education. The high rates of attempted suicide, history of violence and history of drug abuse all furthermore points to the devastating effects bipolar disorder has on individuals and their families. Treatment choice appeared to be a combination of a mood-stabilising agent in combination with an anti-psychotic. It was found that two-thirds of study subjects had consulted with faith- or traditional healers. Significant gender differences appeared in that females were more likely to suffer from comorbid anxiety disorders, have a history of sexual trauma, and be HIV positive whilst men were more likely to have a forensic- and substance-abuse history, experience hallucinations and receive clozapine. Patients presenting with a unipolar manic course of illness, as described in this thesis, may contribute to the search for an etiologically homogeneous sub-group which presents unique phenotype for genetic research and the search for genetic markers in mental illness. A unipolar manic course therefore needs to be considered as a specifier in diagnostic systems in order to heighten the awareness of such a course of illness in bipolar disorder, with a view to future research. / Thesis (MD)--University of Pretoria, 2012. / Psychiatry / unrestricted

BACKGROUND: Concussions are frequent injuries sustained by collegiate athletes. They are often underdiagnosed and may lack adequate treatment due to a lack of standardization in diagnosis and return to play guidelines. Current research suggests that there is an increased risk of subsequent injury after returning to play among previously concussed athletes. Recurrent injuries have a detrimental impact on athletic careers and often result in early retirement from competitive sports. The relationship between sports-related concussions (SRCs) and subsequent musculoskeletal injuries in the NCAA reveals a lack of consensus on return to play (RTP) guidelines and management. Clinical guidelines are evolving with the progressive understanding of SRC pathophysiology. LITERATURE REVIEW FINDINGS: Previous research has found an association between concussion and subsequent acute musculoskeletal injuries, specifically lower extremity injuries. In addition, sports-specific athlete cohorts, such as soccer and football players, have been shown to have a high incidence of SRC. The exact mechanism of action linking SRC and subsequent acute injury is currently unknown. There are numerous etiologic possibilities offered in the literature, including gait alterations, neuromuscular control deficits, and impaired reaction time. However, further research is required to elucidate this relationship—in our review of the current literature we were able to find no studies evaluating the risk of both upper and lower extremity injury following SRC after returning to play in the NCAA population. PROPOSED PROJECT: The goal of the proposed retrospective cohort study of BU athletes is to evaluate the impact of concussion history on recurrent injury after returning to sports. Study inclusion criteria will consist of BU athletes diagnosed with concussions at a single NCAA Division I institution. This cohort will be collected from NCAA injury database and electronic medical records will be matched with controls who have not sustained concussions over the same period. CONCLUSIONS: History of concussion is associated with increased risk of recurrent injury in athletes, specifically at the collegiate level. Further work must be done at all levels of athletics to standardize return to play guidelines to prevent this increased risk of subsequent injury. SIGNIFICANCE: This study will help determine whether a relationship exists between concussion and subsequent sports-related injuries. This information may allow for closer monitoring of return to play and treatment of concussions. Additionally, it may encourage further developments to diagnostic criteria, education, and injury prevention.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20809号 / 医博第4309号 / 新制||医||1025(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 妹尾 浩, 教授 坂井 義治, 教授 戸井 雅和 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM