Lesion detection and classification in breast cancer: evaluation of approaches based on morphological features, tracer kinetic modelling and semi-quantitative parameters in MR functional imaging (DCE-MRI)

The diagnosis, grading and classification of tumours has benefited considerably from the development of DCE-MRI which is now essential to the adequate clinical management of many tumour types due to its capability in detecting active angiogenesis. Several strategies have been proposed for DCE-MRI evaluation. Visual inspection of contrast agent concentration curves vs time is a very simple yet operator dependent procedure, therefore more objective approaches have been developed in order to facilitate comparison between studies. In so called model free approaches, descriptive or heuristic information extracted from time series raw data have been used for tissue classification. The main issue concerning these schemes is that they have not a direct interpretation in terms of physiological properties of the tissues. On the other hand, model based investigations typically involve compartmental tracer kinetic modelling and pixel-by-pixel estimation of kinetic parameters via non-linear regression applied on region of interests opportunely selected by the physician. This approach has the advantage to provide parameters directly related to the pathophysiological properties of the tissue such as vessel permeability, local regional blood flow, extraction fraction, concentration gradient between plasma and extravascular-extracellular space. Anyway, nonlinear modelling is computational demanding and the accuracy of the estimates can be affected by the signal-to-noise ratio and by the initial solutions. The principal aim of this thesis is investigate the use of semi-quantitative and quantitative parameters for segmentation and classification of breast lesion. The objectives can be subdivided as follow: describe the principal techniques to evaluate time intensity curve in DCE-MRI with focus on kinetic model proposed in literature; to evaluate the influence in parametrization choice for a classic bi-compartmental kinetic models; to evaluate the performance of a method for simultaneous tracer kinetic modelling and pixel classification; to evaluate performance of machine learning techniques training for segmentation and classification of breast lesion.

Clinicopathologic findings in thoroughbred horses during a high speed – short term competition

The aim of this study was to examine whether a real high speed-short term competition influences clinicopathological data focusing on muscle enzymes, iron profile and Acute Phase Proteins. 30 Thoroughbred racing horses (15 geldings and 15 females) aged between 4-12 years (mean 7 years), were used for the study. All the animals performed a high speed-short term competition for a total distance of 154 m in about 12 seconds, repeated 8 times, within approximately one hour (Niballo Horse Race). Blood samples were obtained 24 hours before and within 30 minutes after the end of the races. On all samples were performed a complete blood count (CBC), biochemical and haemostatic profiles. The post-race concentrations for the single parameter were corrected using an estimation of the plasma volume contraction according to the individual Alb concentration. Data were analysed with descriptive statistics and the percentage of variation from the baseline values were recorded. Pre- and post-race results were compared with non-parametric statistics (Mann Whitney U test). A difference was considered significant at p<0.05. A significant plasma volume contraction after the race was detected (Hct, Alb; p<0.01). Other relevant findings were increased concentrations of muscular enzymes (CK, LDH; p<0.01), Crt (p<0.01), significant increased uric acid (p<0.01), a significant decrease of haptoglobin (p<0.01) associated to an increase of ferritin concentrations (p<0.01), significant decrease of fibrinogen (p<0.05) accompanied by a non-significant increase of D-Dimers concentrations (p=0.08). This competition produced relevant abnormalities on clinical pathology in galloping horses. This study confirms a significant muscular damage, oxidative stress, intravascular haemolysis and subclinical hemostatic alterations. Further studies are needed to better understand the pathogenesis, the medical relevance and the impact on performance of these alterations in equine sport medicine.

The impact of grants in the agriculture sector in Albania: a counterfactual approach

The objective of this study is to measure the impact of the national subsidy scheme on the olive and fruit sector in two regions of Albania, Shkodra and Fier. From the methodological point of view, we use a non- parametric approach based on the propensity score matching. This method overcomes problem of the missing data, by creating a counterfactual scenario. In the first step, the conditional probability to participate in the program was computed. Afterwards, different matching estimators were applied to establish whether the subsidies have affected sectors performance. One of the strengths of this study stays in the data. Cross-sectional primary data was gathered through about 250 interviews.. We have not found empirical evidence of significant effects of government aid program on production. Differences in production found between beneficiaries and non-beneficiaries disappear after adjustment by the conditional probability of participating into the program. This suggests that subsidized farmers would have performed better than the subsidized households even in the absence of production grants, revealing program self-selection. On the other hand, the scheme has affected positively the farm structure increasing the area under cultivation, but yields has not increased for beneficiaries compared to non beneficiaries. These combined results shed light on the reason of the missed impact. It could be reasonable to believe that the new plantation, in particular in the case of olives, has not yet reached full production. Therefore, we have reasons to believe on positive impacts in the future. Concerning some qualitative results, the extension of area under cultivation is strongly conditioned by the small farm size. This together with a thin land market makes extremely difficult the expansion beyond farm boundaries.

Wavelet estimation in diffusions with periodicity

Wir betrachten einen zeitlich inhomogenen Diffusionsprozess, der durch eine stochastische Differentialgleichung gegeben wird, deren Driftterm ein deterministisches T-periodisches Signal beinhaltet, dessen Periodizität bekannt ist. Dieses Signal sei in einem Besovraum enthalten. Wir schätzen es mit Hilfe eines nichtparametrischen Waveletschätzers. Unser Schätzer ist von einem Wavelet-Dichteschätzer mit Thresholding inspiriert, der 1996 in einem klassischen iid-Modell von Donoho, Johnstone, Kerkyacharian und Picard konstruiert wurde. Unter gewissen Ergodizitätsvoraussetzungen an den Prozess können wir nichtparametrische Konvergenzraten angegeben, die bis auf einen logarithmischen Term den Raten im klassischen iid-Fall entsprechen. Diese Raten werden mit Hilfe von Orakel-Ungleichungen gezeigt, die auf Ergebnissen über Markovketten in diskreter Zeit von Clémencon, 2001, beruhen. Außerdem betrachten wir einen technisch einfacheren Spezialfall und zeigen einige Computersimulationen dieses Schätzers.

Anwendungen des Komplexe-Masse-Renormierungsschemas in effektiver Feldtheorie

Der erste Teil der vorliegenden Dissertation befasst sich mit der Untersuchung der perturbativen Unitarität im Komplexe-Masse-Renormierungsschema (CMS). Zu diesem Zweck wird eine Methode zur Berechnung der Imaginärteile von Einschleifenintegralen mit komplexen Massenparametern vorgestellt, die im Grenzfall stabiler Teilchen auf die herkömmlichen Cutkosky-Formeln führt. Anhand einer Modell-Lagrangedichte für die Wechselwirkung eines schweren Vektorbosons mit einem leichten Fermion wird demonstriert, dass durch Anwendung des CMS die Unitarität der zugrunde liegenden S-Matrix im störungstheoretischen Sinne erfüllt bleibt, sofern die renormierte Kopplungskonstante reell gewählt wird. Der zweite Teil der Arbeit beschäftigt sich mit verschiedenen Anwendungen des CMS in chiraler effektiver Feldtheorie (EFT). Im Einzelnen werden Masse und Breite der Deltaresonanz, die elastischen elektromagnetischen Formfaktoren der Roperresonanz, die elektromagnetischen Formfaktoren des Übergangs vom Nukleon zur Roperresonanz sowie Pion-Nukleon-Streuung und Photo- und Elektropionproduktion für Schwerpunktsenergien im Bereich der Roperresonanz berechnet. Die Wahl passender Renormierungsbedingungen ermöglicht das Aufstellen eines konsistenten chiralen Zählschemas für EFT in Anwesenheit verschiedener resonanter Freiheitsgrade, so dass die aufgeführten Prozesse in Form einer systematischen Entwicklung nach kleinen Parametern untersucht werden können. Die hier erzielten Resultate können für Extrapolationen von entsprechenden Gitter-QCD-Simulationen zum physikalischen Wert der Pionmasse genutzt werden. Deshalb wird neben der Abhängigkeit der Formfaktoren vom quadrierten Impulsübertrag auch die Pionmassenabhängigkeit des magnetischen Moments und der elektromagnetischen Radien der Roperresonanz untersucht. Im Rahmen der Pion-Nukleon-Streuung und der Photo- und Elektropionproduktion werden eine Partialwellenanalyse und eine Multipolzerlegung durchgeführt, wobei die P11-Partialwelle sowie die Multipole M1- und S1- mittels nichtlinearer Regression an empirische Daten angepasst werden.

Essays in Health Economics

In the first chapter we develop a theoretical model investigating food consumption and body weight with a novel assumption regarding human caloric expenditure (i.e. metabolism), in order to investigate why individuals can be rationally trapped in an excessive weight equilibrium and why they struggle to lose weight even when offered incentives for weight-loss. This assumption allows the theoretical model to have multiple equilibria and to provide an explanation for why losing weight is so difficult even in the presence of incentives, without relying on rational addiction, time-inconsistency preferences or bounded rationality. In addition to this result we are able to characterize under which circumstances a temporary incentive can create a persistent weight loss. In the second chapter we investigate the possible contributions that social norms and peer effects had on the spread of obesity. In recent literature peer effects and social norms have been characterized as important pathways for the biological and behavioral spread of body weight, along with decreased food prices and physical activity. We add to this literature by proposing a novel concept of social norm related to what we define as social distortion in weight perception. The theoretical model shows that, in equilibrium, the effect of an increase in peers' weight on i's weight is unrelated to health concerns while it is mainly associated with social concerns. Using regional data from England we prove that such social component is significant in influencing individual weight. In the last chapter we investigate the relationship between body weight and employment probability. Using a semi-parametric regression we show that men and women employment probability do not follow a linear relationship with body mass index (BMI) but rather an inverted U-shaped one, peaking at a BMI way over the clinical threshold for overweight.

Segmentation of brain tumor images based on atlas-registration combined with a Markov-Random-Field lesion growth model

We present an automatic method to segment brain tissues from volumetric MRI brain tumor images. The method is based on non-rigid registration of an average atlas in combination with a biomechanically justified tumor growth model to simulate soft-tissue deformations caused by the tumor mass-effect. The tumor growth model, which is formulated as a mesh-free Markov Random Field energy minimization problem, ensures correspondence between the atlas and the patient image, prior to the registration step. The method is non-parametric, simple and fast compared to other approaches while maintaining similar accuracy. It has been evaluated qualitatively and quantitatively with promising results on eight datasets comprising simulated images and real patient data.

Burden of abdominal obesity in cardiac rehabilitation patients: Results from the Swiss CaRe study

BACKGROUND: The burden of abdominal obesity (AO) and its association with other cardiovascular risk factors is not known in coronary artery disease (CAD) patients attending cardiac rehabilitation (CR). The aim of this study was, therefore, to investigate the prevalence of AO and differences in cardiovascular risk factors between AO and non-AO patients. METHODS: 415 consecutive male CAD patients (mean age 58 ± 11 years) attending a three-month outpatient CR programme were assessed. Differences in cardiovascular risk profile, including blood lipids, psychosocial and socioeconomic status and exercise capacity, were compared in relation to AO and corrected for obesity measured by body-mass index (BMI) in a multivariate analysis. RESULTS: Mean waist circumference was 102 ± 11 cm. Patients of lower educational level had a higher prevalence of AO (p = 0.021) than patients with a higher educational level. AO was significantly associated with diabetes (p = 0.003) and hypertension (p <0.001). In AO patients, HDL-C levels were lower (p <0.001) and triglyceride levels higher (p = 0.006) than in non-AO patients. There was no difference in exercise capacity between AO and non-AO patients, but AO patients had a higher resting heart rate (p = 0.021). CONCLUSION: AO is highly prevalent in CAD patients attending CR. AO is, independently of BMI, associated with metabolic lipid disorders and autonomic cardiovascular dysregulation, suggesting an increased cardiovascular risk. AO patients therefore need particular attention during CR and follow-up care.

Geophysical Analysis of Seasonal Montandon Gravel Ridge Water Table Fluctuation and Moisture Gradient Variation due to Storm Events

The purpose of this research project is to continue exploring the Montandon Long-Term Hydrologic Research Site(LTHR) by using multiple geophysical methods to obtain more accurate and precise information regarding subsurface hydrologic properties of a local gravel ridge,which are important to both the health of surrounding ecosystems and local agriculture. Through using non-invasive geophysical methods such as seismic refraction, Direct Current resistivity and ground penetrating radar (GPR) instead of invasive methods such as boreholedrilling which displace sediment and may alter water flow, data collection is less likely to bias the data itself. In addition to imaging the gravel ridge subsurface, another important researchpurpose is to observe how both water table elevation and the moisture gradient (moisture content of the unsaturated zone) change over a seasonal time period and directly after storm events. The combination of three types of data collection allows the strengths of each method combine together and provide a relatively strongly supported conclusions compared to previous research. Precipitation and geophysical data suggest that an overall increase in precipitation during the summer months causes a sharp decrease in subsurface resistivity within the unsaturated zone. GPR velocity data indicate significant immediate increase in moisture content within the shallow vadose zone (< 1m), suggesting that rain water was infiltrating into the shallow subsurface. Furthermore, the combination of resistivity and GPR results suggest that the decreased resistivity within the shallow layers is due to increased ion content within groundwater. This is unexpected as rainwater is assumed to have a DC resistivity value of 3.33*105 ohm-m. These results may suggest that ions within the sediment must beincorporated into the infiltrating water.

Proteomic changes in the rat brain induced by homogenous irradiation and by the bystander effect resulting from high energy synchrotron X-ray microbeams

Abstract Purpose: To further evaluate the use of microbeam irradiation (MBI) as a potential means of non-invasive brain tumor treatment by investigating the induction of a bystander effect in non-irradiated tissue. Methods: Adult rats were irradiated with 35 or 350 Gy at the European Synchotron Research Facility (ESRF), using homogenous (broad beam) irradiation (HI) or a high energy microbeam delivered to the right brain hemisphere only. The proteome of the frontal lobes were then analyzed using two-dimensional electrophoresis (2-DE) and mass spectrometry. Results: HI resulted in proteomic responses indicative of tumourigenesis; increased albumin, aconitase and triosphosphate isomerase (TPI), and decreased dihydrolipoyldehydrogenase (DLD). The MBI bystander effect proteomic changes were indicative of reactive oxygen species mediated apoptosis; reduced TPI, prohibitin and tubulin and increased glial fibrillary acidic protein (GFAP). These potentially anti-tumourigenic apoptotic proteomic changes are also associated with neurodegeneration. However the bystander effect also increased heat shock protein (HSP) 71 turnover. HSP 71 is known to protect against all of the neurological disorders characterized by the bystander effect proteome changes. Conclusions: These results indicate that the collective interaction of these MBI-induced bystander effect proteins and their mediation by HSP 71, may confer a protective effect which now warrants additional experimental attention.

Who benefits from peer support in psychiatric institutions?

This study examines the influence of recovery-oriented peer events on participants' recovery attitudes and explores who benefits most from such events. Changes in participants' recovery attitudes were evaluated (pre, post, follow-up), and compared with changes of control groups. Distributions of recovery-related values in subgroups were analyzed descriptively. The results of non-parametric tests (Friedman) showed participants with significantly higher values in the dimension Recovery is possible directly after the interventions (P = 0.006), but not 6 months later, and not in comparison with members of control groups. On a descriptive level, women, participants with schizophrenia and with two or more episodes of the disorder showed higher recovery-related values compared to men, participants with an affective disorder and only one episode. Within their feedback, organizations and peers express a positive view of peer support, but evidence for a positive impact of the evaluated peer events on recovery attitude is limited.

A Natural Gas Severance Tax-An Economic Analysis

This study estimates the economic effects of a severance tax on the market for natural gas produced from shale sources using non-conventional extraction methods, such as horizontal drilling and fracking. Results suggest that a severance tax of 5% would increase the price of natural gas by as much as 3.82% and decrease gas extraction by an estimated 1.16% to a value of 9.52%. If applied to the Commonwealth of Pennsylvania in the United States, a 5% severance tax is estimated to raise between US$443 and $486 million per year in public revenue. The marginal deadweight loss associated with a 5% severance tax is estimated between 1.27% and 12.85% of the last dollar earned. The burden of this tax falls on both producers and consumers and depends upon the underlying assumptions made regarding the price responsiveness of consumers and producers. Under plausible assumptions, a family consuming 1000 MMcfs (approximate to 2.8 x 10(4) m(3)) per year of natural gas is estimated to pay an additional $100 per year after the implementation of a 5% severance tax.

Estimating the Size of the Homeless Population in Budapest

Since 1990, the issue of homelessness has become increasingly important in Hungary as a result of economic and structural changes. Various suggestions as to how the problem may be solved have always been preceded by the question "How many homeless people are there?" and there is still no official consensus as to the answer. Counting of the homeless is particularly difficult because of the bias in the initial sampling frame due to two factors that characterise this population: the definition of homelessness, and its 'hidden' nature. David aimed to estimate the size of the homeless population of Budapest by using two non-standard sampling methods: snowball sampling and the capture-recapture method. Her calculations are based on three data sets: one snowball data set and two independent list data sets. These estimators, supported by other statistical data, suggest that in 1999 there were about 8000-10000 homeless people in Budapest.

POOR PERFORMANCE OF BOOTSTRAP CONFIDENCE INTERVALS FOR THE LOCATION OF A QUANTITATIVE TRAIT LOUCS

The aim of many genetic studies is to locate the genomic regions (called quantitative trait loci, QTLs) that contribute to variation in a quantitative trait (such as body weight). Confidence intervals for the locations of QTLs are particularly important for the design of further experiments to identify the gene or genes responsible for the effect. Likelihood support intervals are the most widely used method to obtain confidence intervals for QTL location, but the non-parametric bootstrap has also been recommended. Through extensive computer simulation, we show that bootstrap confidence intervals are poorly behaved and so should not be used in this context. The profile likelihood (or LOD curve) for QTL location has a tendency to peak at genetic markers, and so the distribution of the maximum likelihood estimate (MLE) of QTL location has the unusual feature of point masses at genetic markers; this contributes to the poor behavior of the bootstrap. Likelihood support intervals and approximate Bayes credible intervals, on the other hand, are shown to behave appropriately.

COVARIATE-ADJUSTED NONPARAMETRIC ANALYSIS OF MAGNETIC RESONANCE IMAGES USING MARKOV CHAIN MONTE CARLO

Permutation tests are useful for drawing inferences from imaging data because of their flexibility and ability to capture features of the brain that are difficult to capture parametrically. However, most implementations of permutation tests ignore important confounding covariates. To employ covariate control in a nonparametric setting we have developed a Markov chain Monte Carlo (MCMC) algorithm for conditional permutation testing using propensity scores. We present the first use of this methodology for imaging data. Our MCMC algorithm is an extension of algorithms developed to approximate exact conditional probabilities in contingency tables, logit, and log-linear models. An application of our non-parametric method to remove potential bias due to the observed covariates is presented.