Auf gesellschaftlicher Ebene spielt Bildung in Deutschland eine wichtige Rolle bei der Reproduktion kultureller Normen und Werte. Aber auch auf individueller Ebene fördert sie die soziale Teilhabe und die Identitätsentwicklung. Trotz des Prinzips der Chancengleichheit bleibt der Bildungserfolg in Deutschland nach wie vor stark von der sozialen Herkunft und somit vom familiären Kontext abhängig. In dieser Dissertation wird das Zusammenspiel verschiedener Aspekte des familiären Zusammenlebens und ihr Einfl uss auf die Bildung von Kindern untersucht. Die Analysen basieren auf Daten der Startkohorte 3 (SC3) des Nationalen Bildungspanels (NEPS) sowie des Beziehungs- und Familienpanels (pairfam). Im Fokus stehen Kinder, die mindestens die vierte und höchstens die zehnte Klasse besuchen. Die Ergebnisse zeigen, dass die Qualität der innerfamiliären Beziehungen eine Rolle für die Bildung dieser Kinder spielt. In der deutschen Forschung fand dieses Thema bislang wenig Beachtung, ebenso wie die Einbeziehung der Wahrnehmung und Perspektive der Kinder auf die Familie. Untersucht wird insbesondere, wie Veränderungen innerhalb der Familie - sowohl hinsichtlich ihrer Form als auch der Beziehungsqualität - die Mathematikleistungen und Bildungsaspirationen von Kindern beeinflussen. Ein besonderer Fokus liegt dabei auf dem Vergleich zwischen Kindern, die mit beiden leiblichen Elternteilen in einem Haushalt aufwachsen, und solchen, die dies nicht tun. Die Erkenntnisse dieser Dissertation unterstreichen die zentrale Bedeutung familiärer Beziehungsdynamiken für Bildungsprozesse und betonen, wie wichtig es ist, die Perspektiven und Bedürfnisse der Kinder stärker in den Mittelpunkt zu stellen. Es zeigt sich, dass eine umfassende Förderung, die sowohl emotionale als auch materielle Unterstützung umfasst, unerlässlich ist, um negative Einflüsse auf schulische Leistungen, insbesondere in instabilen Familienverhältnissen, zu vermeiden. Die Ergebnisse bieten wertvolle Impulse für die Familien- und Bildungspolitik und verdeutlichen die Notwendigkeit ganzheitlicher Unterstützungsangebote, die allen Kindern gleiche Bildungs- und Entwicklungschancen ermöglichen, unabhängig von ihrer familiären Situation. Dabei ist es von zentraler Bedeutung, Programme zu entwickeln, die die emotionalen, sozialen und akademischen Bedürfnisse der Kinder gleichermaßen berücksichtigen.

The overall objectives of this thesis were (1) to evaluate different methods of serum collection for the detection of FTPI via refractometry in neonatal calves, (2) to evaluate a pyroelectric infrared monitoring system assessing movement of preweaned calves, and (3) to assess temperature preferences for heated calf hutches during winter in newborn dairy calves.

The on-farm monitoring of FTPI to optimize colostrum management is an integral part of heifer rearing. Therefore, the objective of the first study was to compare 4 different methods of serum collection to assess failed transfer of passive immunity (FTPI) in dairy calves. We hypothesized that centrifuged serum, filtered serum and clotted serum at room temperature, and clotted serum at refrigerator temperature measured with Brix refractometry would highly correlate with IgG concentration assessed by radial immunodiffusion (RID; gold standard) in centrifuged serum. Blood samples were collected from 321 newborn dairy calves. In centrifuged serum, serum clotted at room temperature, serum clotted at refrigerator temperature, and filtered serum, total solids (TS) in % Brix, and IgG concentrations measured with RID were highly correlated. Regarding the refractometry results among the different serum types, the TS of clotted and filtered serum showed high correlation coefficients compared with the TS results of centrifuged serum. Filtered serum was slightly less accurate. All serum types are valid methods to detect an FTPI in dairy calves, if the specific Brix thresholds for each serum type are considered. Nevertheless, serum clotted at refrigerator temperature should not be the preferred method to avoid the risk of hemolysis.

Sensor technology is becoming more and more popular in calf rearing, and could provide important information about health and performance in the future. The objective of this study was to correlate movement assessed by a pyroelectric infrared sensor system (IMS) in preweaned dairy calves with lying and standing time assessed by a 3D accelerometer considering the temperature-humidity index (THI). A total of 35 dairy calves (1–7 d of age) were enrolled in the study and 20 calves were included in the final analyses. Number of movements per hour measured by the IMS was compared with data obtained with a validated 3D accelerometer. The Pearson correlation coefficient between both standing and lying time and the number of movements was r = 0.85 and r = −0.85, respectively. The Pearson correlation coefficients were only slightly influenced by THI. Our data show that the number of movements of dairy calves measured by IMS were highly correlated with the chosen gold standard reference method. High THI slightly affects the measurement accuracy of IMS. Newborn calves face challenges adapting from the warm uterine environment to cold ambient temperatures, often experiencing cold stress. Therefore, the objective of the third study was to assess the preferred inside hutch temperature of dairy calves in their first week after birth during Central European winter and to compare lying behavior in heated and non-heated hutches. Calves spent 90% of their time inside a hutch, irrespective of the number of heat lamps switched on. No preference for a specific inside hutch temperature existed. There was, however, a tendency for the location of the hutches closest to the control calf. Lying behavior did not differ between preference and control phase, but decreased marginally with increasing age. These findings could indicate that there is a need of shelter for calves in winter, but without a preference for a certain inside hutch temperature under the prevailing conditions of deep straw bedding, highly sufficient milk supply and an ambient temperature of 5.4 ± 3.3°C.

Overall, the results of this thesis show that (1) there are different valuable methods to harvest serum for the detection of FTPI in newborn dairy calves, even without centrifugation, (2) there is a high correlation between the assessment of movement with an IMS and the lying and standing times assessed with a 3D accelerometer data logger, and (3) there is no preference for a certain inside hutch temperature in mild winters, but a preference for social contact. All three studies provide results that can contribute to concrete decision-making and practical application for good calf health monitoring

This dissertation consists of four independent chapters that contribute to the literature of labor and public economics. Institutions influence the labor market outcomes and redistribution of incomes, ultimately shaping the income distributions we observe today. The first chapter focuses on the legal framework for migrants on the German labor market and highlights the differences between two groups of foreign workers: citizens of EU member states and citizens of third countries. Germany’s institutions favor EU citizens in terms of mobility, leading to higher wages and higher unemployment. The second and third chapter take a broader perspective on the entire German population and how the earnings dynamics and income inequalities differ between birth cohorts – including welfare implications. The final chapter examines the German tax and transfer system and its implicit welfare weights along the income distribution. Chapter 1 examines the difference in mobility on the German labor market for citizens of EU member states and third countries and its consequences. Immigrants from non-EU countries face considerably higher barriers. I develop a wage posting model to illustrate the underlying mechanisms and provide reduced-form evidence of the causal effects of EU citizenship on labor market outcomes. I exploit the 2004 enlargement of the European Union, which granted EU status to immigrants from ten Eastern European countries residing in Germany. Using a difference-in-differences framework as well as an event-study and rich administrative data from the Sample of Integrated Labour Market Biographies (SIAB), I compare the labor market trajectories of this group before and after enlargement with those of non-EU immigrants who did not benefit from such a change in legal status. The results show that EU citizenship increases wages while unemployment rates increase. These effects are persistent over time. Chapter 2 puts a spotlight on the earnings dynamics of different birth cohorts in Germany. Women born later experience greater earnings growth volatility at given ages than older cohorts. This implies a welfare loss due to increased earnings risk. However, German registry data for the years 2001-2016 reveal a moderation in higher-order earnings risk: Men and women born later face higher skewness in earnings changes, indicating fewer large decreases than increases, and lower kurtosis at younger ages, implying fewer large earnings changes. These trends point at a welfare increase and remain true for 5-year earnings changes, which are more reflective of persistent changes. During the Great Recession, males' skewness dropped sharply; younger women were unaffected. Chapter 3 examines the welfare effects of changes in real income distributions between German birth cohorts. We calculate the relative income increase at every income level of the cohort born earlier necessary to make a person behind the veil of ignorance indifferent between being born into either cohort. With relatively low risk aversion, individuals born in the 1970s are on average better off with respect to labor earnings than those born in the 1960s and individuals born in the 1980s enjoy higher welfare than those born in the 1970s. At higher levels of risk aversion, the welfare gain diminishes, implying that increased earnings inequality offset much of the gain of increased average earnings. However, redistribution through households and the welfare state cushioned much of the increase in inequality. Focusing on post-government household income instead of individual labor earnings, for both women and men, the proportional income increase for the earlier cohort for someone to be indifferent between being born in the earlier or the later cohort is substantially larger, especially at higher degrees of relative risk aversion. Chapter 4 investigates the question: How much does society value redistribution? The common method to derive inverse-optimum welfare weights is by inverting an optimal-tax model. Our alternative imposes fewer restrictions on labor supply and enables comparisons across household types. We use a structural labor supply model to calculate the marginal value of public funds for various small tax reductions, directly linked to welfare weights. An application to Germany finds: i) The tax-transfer system is optimal if society values one additional Euro for the bottom decile three times as much as for the median. ii) At low-medium incomes, weights for couples exceed those for singles substantially.

Mit der tiefen Hirnstimulation (THS) steht für pharmakorefraktäre Fälle oder in fortgeschrittenen Erkrankungsstadien von Bewegungsstörungen ein effektives und etabliertes neuromodulatorisches Therapieverfahren zur Verfügung. Durch Verwendung moderner multimodaler bildgebender Verfahren in Kombination mit Aufzeichnungen von neuronaler Aktivität erweiterte sich das Verständnis des Wirkmechanismus dieser Therapie von der fokalen Modulation pathologischer neuronaler Aktivitätsmuster um Netzwerkeffekte in weiteren Hirnregionen, die mono- oder polysynaptisch mit den stimulierten Arealen in Verbindung stehen. Die klinischen Effekte werden in einem Großteil der hierzu existierenden Publikationen über semiquantitative Scores und Skalen erfasst, die oftmals für die Patient:innen relevante spezifische Symptomänderungen nur unzureichend abbilden. Eine präzise und dynamische, quantitative Erfassung der individuellen Symptomschwere von Patient:innen mit Bewegungsstörungen ist erforderlich, um Netzwerkstrukturen zu differenzieren, deren Stimulation vorteilhaft mit der Modulation von einzelnen Zielsymptomen assoziiert ist. Unter Nutzung der technischen Möglichkeiten moderner Stimulationselektroden, die eine fokussierte Ausrichtung des elektrischen Feldes ermöglichen, lassen sich hiervon personalisierte Optimierungen der Stimulationsparameter ableiten. Die vorgelegten Arbeiten beschreiben die Verwendung der quantitativen, sensorgestützten Bewegungsanalyse zur Charakterisierung von pathologischen Veränderungen des Gangbildes bei Patient:innen mit tiefer Hirnstimulation mit dem Ziel, aus diesen Funktionsmarker abzuleiten, die diagnostisch sowie zur Beurteilung und Optimierung von Therapien herangezogen werden können. Die in dieser Habilitationsschrift zusammengefassten Arbeiten widmen sich der Optimierung eines etablierten neuromodulatorischen Therapieverfahrens mit dem Ziel der besseren Behandlung von Gangstörungen, die als Symptom einer Bewegungsstörung oder als unerwünschte Nebenwirkung der tiefen Hirnstimulation auftreten. Die Verwendung quantitativer Bewegungsanalysen zusätzlich zu klinischen Skalen und Scores ermöglichten therapeutische Interventionen hinsichtlich ihrer Auswirkungen auf die Zielsymptome präzise zu charakterisieren und klinische Effekte in Bezug zu neuroanatomischen Strukturen und Netzwerken zu setzen. Darüber hinaus wurde der Einfluss der räumlichen Untersuchungsbedingungen auf diagnostisch und therapeutisch relevante Merkmale von Gangstörungen in diesem Kontext evaluiert. Die in dieser Schrift vorgestellten Arbeiten belegen somit die Relevanz quantitativer Bewegungsanalysen bei der Optimierung und Weiterentwicklung der tiefen Hirnstimulation hin zu einer symptomspezifischen und personalisierten Therapieform.  

Die vorliegende Habilitationsarbeit untersucht die Rolle der tiefen Hirnstimulation (THS) bei der Behandlung des Morbus Parkinson und deren potenzielle Anwendung zur Linderung chronischer Rückenschmerzen, einer häufigen und stark belastenden Begleiterkrankung bei an Morbus Parkinson erkrankten Patientinnen und Patienten. Seit der Einführung der THS in den späten 1980er-Jahren hat diese Methode die Therapie motorischer Symptome des Morbus Parkinson revoluoniert. Die positve Wirkung der THS auf die Hauptsymptome Rigor, Tremor und Brady/Akinese ist gut belegt, jedoch bleibt ihre Wirkung auf andere, stark belastende Symptome wie axiale Instabilität, Sprachprobleme und chronische Rückenschmerzen eingeschränkt und wenig erforscht. Chronische Rückenschmerzen sind bei Parkinson-Patientinnen und -Patenten eine häufige, jedoch unterrepräsentierte Komorbidität. Sie stehen in Zusammenhang mit spinalen Deformitäten wie der Camptocormie und dem Pisa-Syndrom, deren pathophysiologische Mechanismen nicht vollständig verstanden sind. In dieser Arbeit wird untersucht, inwiefern die THS durch die Modulaton neuronaler Netzwerke zur Linderung von Rückenschmerzen beitragen könnte. Die Erweiterung der Indikation der THS auf Rückenschmerzen könnte langfristig zu einer deutlichen Verbesserung der Lebensqualität der Patientinnen und Patenten führen und eine therapeutische Alternative gegenüber komplikationsbehafteter Wirbelsäulenoperationen darstellen. Die Arbeit befasst sich zudem mit den erhöhten Komplikationsraten und biomechanischen Herausforderungen bei Wirbelsäuleneingriffen an Parkinson- Patentinnen und -Patienten, die aufgrund ihrer neurodegenerativen Erkrankung besondere Anforderungen an die operative Therapie stellen. Studien zeigen eine überdurchschnittlich hohe Komplikatonsrate bei stabilisierenden Eingriffen an der Wirbelsäule, was eine detaillierte Analyse der Risikofaktoren und eine angepasste operative Strategie erforderlich macht. Eine neuartige Methode zur dynamischen, nicht-invasiven Bewertung der biomechanischen Eigenschaften der Wirbelsäule wird vorgestellt, um mögliche Fehlbelastungen und Risikofaktoren präziser zu identifizieren und präventive Ansätze zu entwickeln. Schließlich thematisiert die Arbeit THS-assoziierte Komplikationen, insbesondere Implantat-assoziierte Infektionen, die ein relevantes Risiko darstellen. Die Arbeit schlägt optimierte Strategien zur Prävention und Behandlung dieser Infektionen, einschließlich durch niedrigvirulente Erreger, vor. Zusammengefasst bietet diese Habilitationsarbeit neue Einblicke in die therapeutischen Potenziale und Herausforderungen der THS für die komplexe Symptomatik des Morbus Parkinson. Die Ergebnisse dieser Arbeit könnten als Basis für eine Erweiterung der Indikationsfelder der THS und die Entwicklung interdisziplinärer, individualisierter Behandlungsansätze dienen, um die Lebensqualität und das Wohlbefinden von Patentinnen und Patienten mit Morbus Parkinson nachhaltig zu verbessern.

In der vorliegenden Habilitationsschrift wurde mittels bildgebender Verfahren die Ausprägung kardiovaskulärer und zerebrovaskulärer Schäden im Zusammenhang mit eingeschränkter Nierenfunktion untersucht. Zusätzlich wurden modifizierbare Risikofaktoren für Organschäden identifiziert, die durch entsprechende Kontrolle die Prognose verbessern können. In der kardialen MRT wurden bei fast der Hälfte der untersuchten nierenkranken Patienten strukturelle Schäden des linken Ventrikels, wie Hypertrophie, Dilatation und konzentrisches Remodeling, nachgewiesen. Modifizierbare Risikofaktoren für diese Schäden waren ein erhöhter BMI, arterielle Hypertonie in der 24 Stunden Blutdruckmessung sowie eine erhöhte Hämoglobin Konzentration im Serum. Anhand der 24 Stunden Blutdruckmessungen konnte gezeigt werden, dass etwa ein Drittel der Patienten mit Einzelmessungen des Blutdrucks fehlklassifiziert wurde, wobei hier insbesondere maskierter Bluthochdruck (in der Arztpraxis wird fälschlich ein normaler Blutdruck gemessen) und Weißkittelhypertonie (in der Arztpraxis wird fälschlich ein erhöhter Blutdruck gemessen) von Bedeutung sind. Beide Formen von Fehlklassifikation waren mit einer linksventrikulären Hypertrophie assoziiert. Patienten mit maskiertem Bluthochdruck hatten zusätzlich eine erhöhte Pulswellengeschwindigkeit und einen erhöhten zentralen Augmentationsindex. Diese Ergebnisse unterstreichen die Bedeutung der 24 Stunden Blutdruckmessung zur Vermeidung der Fehlklassifikation von arterieller Hypertonie. Doch nicht nur das Herz-Kreislaufsystem ist bei nierenkranken Patienten gefährdet. In einer weiteren Arbeit konnte gezeigt werden, dass eine eingeschränkte Nierenfunktion auch mit einem erhöhten Demenzrisiko assoziiert ist, sowohl im mittleren als auch im fortgeschrittenen Lebensalter. Dies galt sowohl für akute als auch für chronische Einschränkungen der Nierenfunktion. Diese Assoziation war stärker ausgeprägt, wenn zur Abschätzung der Nierenfunktion der von der Muskelmasse unabhängige Laborparameter Cystatin C verwendet wurde, was bei älteren Patienten mit geringerer Muskelmasse sinnvoll erscheint. Zusätzlich konnte gezeigt werden, dass eine eingeschränkte Nierenfunktion mit zahlreichen in der MRT sichtbaren Gehirnschäden einhergeht, wie lakunären Infarkten, Mikroblutungen sowie einer gestörten mikrostrukturellen Integrität der weißen Hirnsubstanz. In der Volumenmessung verschiedener Gehirnareale zeigte sich abhängig von der Nierenfunktion eine globale Hirnatrophie, jedoch keine selektive Atrophie der Areale, die typischerweise bei Alzheimer-Demenz auftreten, sodass diese Erkrankung als pathogenetische Ursache unwahrscheinlich erscheint. Insgesamt gehen sowohl akute als auch chronische Nierenkrankheiten häufig mit vaskulären Schäden des Herzens und des Gehirns einher. Durch Maßnahmen zur Vermeidung von akuten Nierenschädigungen und Verlangsamung des Progresses einer chronischen Nierenkrankheit sowie Kontrolle weiterer modifizierbarer Risikofaktoren könnte die Gesamtprognose verbessert werden.

Rift Valley fever (RVF) is an emerging mosquito-borne zoonotic disease caused by the Rift Valley Fever virus (RVFV). It heavily impacts human and animal health and the economies of countries where it occurs, especially in Africa, by causing large epizootics, resulting in high morbidity and mortality among livestock and humans. In March 2016, Uganda reported an outbreak of RVF for the first time since 1968 when human cases were recorded in the central region. Since 2016, Uganda has experienced repeated outbreaks of RVF. The disease can cause mass abortions and high calf mortality in animals and flu-like symptoms in humans, occasionally developing into deadly haemorrhagic fever or inflammation of the brain. Humans are likely infected when handling infected livestock or via bites of infected mosquitos. RVF has socio-economic consequences such as bans on international animal trade, disruption of food security, and economic losses for livestock keepers as well as for governments. The design and implementation of control strategies for RVF in Uganda are currently hampered by various limitations, including limited knowledge on the distribution of RVFV and the associated risk factors, gaps in understanding of the local community knowledge on RVF identification, risk factors for disease occurrence, and pathways for RVFV introduction, exposure, and outbreak consequences in humans and animals. While previous seroprevalence studies that have been implemented in Uganda have illustrated the burden and risk factors for RVF in various locations, there is a need to map the distribution of the disease risk across the country. In addition, in Uganda, during RVF outbreaks, control measures have mainly revolved around limited surveillance and preparing situational reports with diminutive inclusion of local community knowledge on predisposing risk factors and disease identification.

To better understand the distribution of RVFV in livestock and the associated risk factors and to develop a risk map, we collected blood samples from cattle, sheep, and goats in a national cross-sectional survey. Secondary data used included information on RVF outbreaks in livestock and humans between 2016-2021 that were obtained from Ugandan Health authorities, and data on climate and other environmental factors that facilitate mosquito breeding from various online databases. Screening of animal samples for exposure to RVFV provided data for generating a RVFV prevalence map. Data obtained were analysed using a Bayesian model which investigated animal-level factors (age, sex, species) and multiple environmental data including meteorological factors, soil types, and altitude. The overall RVFV seroprevalence was 11.3% (95% confidence interval: 10.2–12.3%). Analysis showed that antibodies were more often detected in older animals compared to the young, and more in cattle compared to sheep and goats. The presence of antibodies was also more pronounced in areas with lower precipitation seasonality, haplic planosols (soil types), and lower cattle density. The risk map generated shows that RVF is endemic in several regions including those in northeastern Uganda where outbreaks have not yet been reported. Considering that a district is the smallest administrative unit that has the mandate to design and administer public and animal health interventions at the sub-national level, districts were classified into one of the three RVF risk levels: high, moderate, and low.

To assess pastoralists’ and agropastoralists’ knowledge of RVF, participatory epidemiological studies using various PE techniques were conducted with livestock keepers and key informants in Napak, Lyantonde, Isingiro, and Butebo districts, located in the Ugandan cattle corridor. Napak and Lyantonde districts were classified as pastoralists while Isingiro and Butebo as agropastoralists. Results of semi-structured interviews revealed that livestock keepers in all four districts possessed knowledge of RVF, local names or descriptions for it, and epidemiological risk factors such as the presence of infected mosquitoes, living in flood-prone areas, and excessive rainfall. Pastoralists called RVF Lonyang, which symbolized a disease associated with jaundice, high fever, abortions in pregnant cows, and sudden death in calves. In Butebo, Lyantonde, and Isingiro, participants likened RVF to an Ebola-like disease, referring to haemorrhaging symptoms such as vomiting, blood, bloody stool, and gum bleeding in humans, and abortion in livestock. RVF was ranked among the top ten most important cattle diseases.

Three risk pathways for RVF introduction, exposure, and consequence were assessed by key informants who included animal health and public health workers. For RVFV introduction in an area, infected domestic animals were considered the most important, followed by livestock movement and trade, while infected wild animals were considered the least important. The findings of the practices that promote human exposure to RVF virus were via drinking raw animal blood and milk, handling of infected animal tissues and fluids at parturition, handling of infected meat and fluids at slaughter, and contact with infected domestic animals in the herd. Drinking raw blood and milk was perceived as the highest exposure pathway. The likelihood of consequences ranked highest were high treatment costs, abortions, reduced production, and morbidity.

The findings of this study suggest low-level endemic circulation of RVFV among livestock across Uganda notwithstanding the absence of clinical cases, with cattle as the species with highest exposure. The study further suggests that the risk of RVF is higher in areas with stable precipitation patterns, as this can maintain critical mosquito population levels by replenishing breeding sites. Coupled with precipitation seasonality, the study suggests that RVF risk is higher in areas with planosols, a type of soil with clayey properties that are prone to flooding when inundated, a critical factor in the sustenance of mosquito breeding sites. Further, as RVF is zoonotic, increased numbers of animal exposure could present an increased risk of spillover infection to humans over time.

Practical implications of these findings include targeted RVF surveillance as well as implementation of risk-based control measures. The RVF risk map would guide the Ugandan authorities in the identification of RVF hotspots as well as design and implement appropriate interventions. For instance, it can guide a vaccination strategy to minimize the burden of the disease in livestock and pre-empty zoonotic spillover to humans by implementing periodic and reactive vaccination in high and medium-risk areas respectively. Further, the findings indicate that veterinary authorities need to be vigilant in their surveillance as some of the clinical cases in some areas may have been underreported or not at all. A stronger collaboration between sectors, guided by the One Health framework, would enhance the detection of any new cases especially if data from human and animal surveillance activities are integrated and analysed jointly. The results of the PE study could be integrated into an early warning system and strategies for veterinary and public health service delivery in Uganda.

CLCA4 ist ein Mitglied der CLCA-Genfamilie, das einer tierartspezifischen Genduplikation unterliegt und vorwiegend im Darmtrakt exprimiert wird. CLCA4 scheint eine bislang unbekannte Rolle bei entzündlichen Darmerkrankungen des Menschen wie UC oder CD und bei kolorektalen Tumoren zu spielen. Diese Erkrankungen zeigen in westlichen Gesellschaften eine steigende Prävalenz und verlaufen häufig tödlich. Die Entwicklung diagnostischer und therapeutischer Ansätze zur Behandlung dieser Krankheiten ist ein zentrales Ziel der Forschung. Die Maus, welche ein wichtiges Modelltier für die Erkrankungen darstellt, hat im Gegensatz zum Menschen zwei funktionelle Clca4 Homologe. Die speziesspezifischen Unterschiede in der Genexpression sowohl im gesunden Darm wie auch im pathologisch veränderten Darm sind noch weitegehend unbekannt.

Ziel dieser Arbeit war die Erstellung eines detaillierten Expressionsprofils der murinen und humanen CLCA4-Homologe im Darmtrakt mittels RT-qPCR und ISH. Dabei wurde zunächst die Expression der murinen Vertreter in verschiedenen Abschnitten des Darms sowie entlang der Zotten-Krypten-Achse unter gesunden Bedingungen untersucht. Zusätzlich wurde eine Expressionsanalyse der beiden murinen Vertreter während der frühen postnatalen Entwicklung durchgeführt. Weiterhin wurde deren Expression im Darm eines DSS-Kolitismodells und AOM/DSS-induzierten Darmtumormodells untersucht. Vergleichend wurde die zelluläre Expression des humanen Homologen in gesunden Darmabschnitten und in CRC untersucht.

Die Expression der murinen Clca4 Vertreter zeigte einige Unterschiede. Während Clca4a in Enterozyten des gesamten Darmtraktes exprimiert war, fand sich Clca4b nur in Enterozyten des Dünndarms. Clca4a wurde ausschließlich in den luminalen Bereichen der Schleimhaut nachgewiesen, wohingegen die Expression von Clca4b tiefer reichte. Die in der Literatur beschriebene Expression von humanem CLCA4 im luminalen Epithel des Dickdarms konnte bestätigt werden. Es fand sich aber keine Expression in Enterozyten des Dünndarms. Die murinen und der humane CLCA4 Vertreter zeigten somit zum Teil ein überlappendes, zum Teil unterschiedliches Expressionsmuster.

Die postnatale Expressionsanalyse zeigte, dass Clca4b vor allem in sich noch entwickelnden Mausdärmen exprimiert war, während Clca4a seine höchsten Expressionswerte erst bei adulten Tieren erreichte.

Im DSS-Kolitismodell zeigte sich eine im Vergleich zu Kontrolltieren erhöhte Expression von Clca4a in distalen, erkrankten Dickdarmabschnitten sowie eine extrem starke de-novo-Expression von Clca4b im gesamten entzündeten Kolon. Beide Homologe wiesen besonders starke Signale in der unmittelbaren Nähe von Ulzerationen. Das unterschiedliche Expressionsmuster im gesunden Mausdarm könnte auf verschiedene Funktionen der beiden murinen Clca4 hindeuten. Das Expressionsprofil insbesondere von Clca4b wäre mit der Rolle eines intestinalen Differenzierungsproteins vereinbar.

Im murinen intestinalen Tumormodell wurden beide murinen Clca4 zwar fast gar nicht in den neoplastisch veränderten Darmzellen nachgewiesen, aber massiv in angrenzenden, nicht tumorösen Enterozyten. Dieses Expressionsprofil zeigte auch das CLCA4 in Proben von humanem CRC. Ein solches Expressionsmuster wäre auch mit einer Funktion als Differenzierungsprotein im Darm vereinbar.

Auch wenn sich einige Aspekte der murinen intestinalen Clca4-Expression in der Expression des humanen CLCA4 widerspiegeln, müssen insbesondere die Unterschiede in den Expressionsmustern bei der Interpretation der Ergebnisse aus der translationalen Forschung berücksichtigt und kritisch hinterfragt werden. Die Entdeckung solcher Unterschiede stellt einen zentralen Aspekt dar, die Übertragbarkeit der Ergebnisse vom Tier auf den Menschen besser einzuschätzen und mögliche Limitationen zu erkennen.

Antibiotic resistance is a highly relevant issue today. In the environmental context, chicken manure can serve as a significant source of antibiotic resistance genes, as well as ESBL‑producing and fluoroquinolone-resistant E. coli. Anaerobic digestion, meanwhile, is a promising technology for biogas production and represents an alternative energy source. In this study, antibiotic resistance was monitored during the anaerobic digestion of chicken manure.

In the first study, we monitored the concentration of total E. coli, including ESBL-producing and fluoroquinolone resistant E. coli, during anaerobic digestion in the mesophilic temperature range, both with and without the addition of sawdust as a carbon source. Temperature had a major impact on E. coli concentrations, as well as on the overall anaerobic digestion process and biogas production. While optimizing the C/N ratio by adding sawdust had no significant effect on E. coli concentrations, it did influence nitrogen content during digestion and affected biogas content.

In the second study, we investigated the presence and abundance of antibiotic resistance genes during mesophilic anaerobic digestion of chicken manure. Neither temperature nor the addition of sawdust had a significant impact on ARG abundance.

Additionally, in both studies, chicken manure was analyzed for AMR E. coli and antibiotic resistance genes, confirming that manure is a potential source for the spread of antibiotic resistance in the environment.

To improve the reduction of antibiotic-resistant E. coli during anaerobic digestion, pretreatment methods and co-digestion with sewage sludge have been suggested. While no effective approach was identified for significantly reducing antibiotic resistance genes, pre‑treatment and co-digestion with sewage sludge may still have a positive impact. The persistence of ARGs is likely related to the microbial community composition, which may help maintain ARG levels.

We conclude that anaerobic digestion is an effective process for reducing AMR E. coli; however, our experimental batch setup was not effective for reducing ARGs, indicating the need for further optimization and modification of the process.

Kinetose stellt ein ernst zu nehmendes Problem in der heutigen Zeit dar insbesondere im Hinblick auf autonomes Fahren. Daher müssen die Ursachen von Kinetose zur Vermeidung und für mögliche Therapieansätze ergründet werden. In dieser Arbeit werden zwei Möglichkeiten zur Entstehung von Kinetose experimentell untersucht: die Theorie der Otolithenorganasymmetrie, die aufgrund ungleich verteilter Massen zu Kinetose führen soll, und als neuartiger Ansatz die Hypothese eines Zusammenhangs zwischen intestinalem Mikrobiom und Kinetose im Zusammenhang mit der Darm-Hirn-Achse. Die Messungen für die Berechnung der Otolithenorganasymmetrie beinhalteten VEMP-Messungen, unilaterale Rotation, SVV, OOR-Bestimmung und SVMT. Für den SVMT wurde zusätzlich eine Abfrage des Kinetosezustands auf einer Skala von 1 bis 10 erfragt (MISC). Zur Berechnung der Alpha- und Beta-Diversität sowie zur Prüfung eines Zusammenhangs zwischen Darmflora und Abbruchzeit beim SVMT wurden Stuhlproben der Probanden an die Firma BIOMES geschickt und mittels NGS analysiert. Die Gesamtkohorte umfasste 125 Probanden (51 Frauen und 36 Männer), von denen 87 Datensätze in die Auswertung eingeflossen sind. Anhand der Abbruchzeit des SVMT wurden die Gruppen in sensibel (< 8 Minuten), medium (8 – 16 Minuten) und unsensibel (> 16 Minuten) eingeteilt. Bei höherer Asymmetrie im Sacculus bzw. Utriculus zeigten sich verkürzte Abbruchzeiten, insbesondere bei Frauen. In der Mikrobiomanalyse zeigten sich größtenteils negative Korrelationen zwischen der Anzahl an Mikroben (gemessen in OTU) und der Abbruchzeit. Das heißt, ein geringes Vorkommen führt zu späteren Abbrüchen. Die Untersuchungen zur Otolithenorganasymmetrie lassen weiterhin Fragen offen. Ob die Asymmetrie aufgrund von Massenveränderungen an den Calcit-Kristallen zustande kommt oder der Einfluss von kurzkettigen Fettsäuren und Transmittern die Ursache für die asymmetrische Signalweiterleitung ist, muss in weiterführenden Studien geklärt werden. Mit den Untersuchungen zum Zusammenhang zwischen Mikrobiom und Kinetose ist ein neues Forschungsfeld eröffnet. Die Ergebnisse lassen einen Zusammenhang vermuten, der in weiteren Forschungsarbeiten diskutiert werden sollte.

This thesis explores how LGBTQ+ identities, discourses, and forms of activism are negotiated on social media in Hungary’s increasingly illiberal political environment. As the Hungarian government promotes anti-gender ideology and frames LGBTQ+ existence as a threat to national values, queer lives become sites of ideological contestation. Social media platforms, while implicated in the reproduction of dominant ideologies, also offer alternative spaces for expression, connection, and resistance. The three empirical studies look at the relationship between social media and the socio-politics of Hungary in the experiences of LGBTQ+ youth. Study 1 examines how public discourse on Facebook reflects and reinforces broader ideological tensions around LGBTQ+ issues. By analysing comment sections of online news articles discussing the children’s book Fairyland is for Everyone, the study shows how users draw on discourses of heteronormative familialism, neocolonialism, and diversity and inclusion to either legitimise or delegitimise queerness. The findings highlight how Facebook becomes a contested space where political rhetoric and cultural narratives are both reproduced and challenged in everyday discourse. Study 2 focuses on the identity negotiation processes of LGBTQ+ youth in Hungary and examines how social media affordances shape their experiences. Drawing on interviews with 16 young LGBTQ+ individuals, the study identifies five key social media affordances: anonymity, audience curation, content curation, visibility, and connectivity. These affordances help self-expression and community building in an unsupportive offline environment. However, they also expose users to surveillance and anti-LGBTQ+ content, requiring young users to develop strategies for managing risks and navigating platform dynamics. Study 3 explores how LGBTQ+ activism takes shape on TikTok, a platform that blends personal branding, entertainment, and political commentary. Through a critical technocultural discourse analysis of content from four prominent Hungarian LGBTQ+ creators, the study identifies four creator roles: the Influencer, the Activist, the Influencer- Activist, and the Casual Creator. These roles reflect different strategies for balancing visibility, advocacy, and platform risk within a restrictive political and media landscape. Together, the studies show that LGBTQ+ digital participation in Hungary is a form of adaptive resistance, shaped by platform dynamics and broader ideological pressures. The thesis positions queer engagement with social media as a site of negotiation, conflict, and resilience in the face of illiberal governance.

The intestinal microbiome has a central role in the host's health. In pig production, newborn and weaned piglets are particularly affected by diarrheal diseases of multifactorial origin, which can often only be controlled with antibiotics. However, in line with the “One Health” approach, the aim is to reduce the use of antibiotics in animal husbandry. Targeted modulation of the intestinal microbiome could enable new preventive measures, with the mother sow playing a key role in the initial microbial colonization of newborn piglets and intestinal colonization up to weaning. This approach was investigated by the “OptiBiom” project by analyzing comprehensive microbiome and farm data and combined with targeted microbiome modulation and vaccines to develop new anti-infection strategies. For this purpose, fecal samples were collected from sows before and after birth and their piglets before and after weaning from 20 commercial farms. This thesis analyzes data on the fecal microbiome, antibiotic resistance genes, and bacterial virulence factors. In addition, the use of antibiotics in the corresponding production cycles was documented and evaluated for each farm. The hypothesis that the fecal microbiota composition and the occurrence of resistance and virulence genes differ significantly between the farms and age groups of the animals was investigated. In addition, correlations were examined between farm differences and the sow-piglet relationship, as well as between antibiotic use and the occurrence of bacterial genes. In the first publication, 802 DNA extracts from fecal samples were sequenced using Illumina NextSeq and analyzed for diversity and microbial composition. Detected genera were compared between the different age groups of the animals, and the differences could be visualized using non-metric multidimensional scaling and hierarchical clustering. Data from sows and their piglets were pooled to investigate the family effect on farm differences. In the second publication, quantitative real-time PCR was used to detect resistance genes of relevant antibiotics (sulfonamides, trimethoprim, colistin, cefotaxime) and an integron responsible for spreading resistance genes (Int1). In addition, the samples were examined for relevant pathogens associated with diarrhea in piglets (Clostridioides difficile, Clostridium perfringens, Salmonella, Escherichia/Shigella/Hafnia, Escherichia coli). The farms were analyzed for cluster formation based on their antibiotic use. Latent class analysis was then used to identify the genes relevant for cluster formation in each animal group. Adapted regression models were used to test correlations with the antibiotics administered. The phyla Bacillota (formerly Firmucutes) and Bacteroidota (formerly Bacteroidetes) dominate all samples with varying relative abundances depending on the production phase. Parturition (sows) and weaning (piglets) led to significant changes in the fecal microbiota composition. After birth, the relative abundance of non-dominant genera, in particular, decreased significantly, while previously dominant genera continued to predominate. In piglets, on the other hand, originally non-dominant genera became predominant after weaning, while previously dominant genera from the suckling phase decreased. The microbiome of suckling piglets was the most uniform and showed the least diversity. Combining the microbiome data of sows and their piglets revealed a distinctive cluster formation for two farms. The use of antibiotics on each farm led to different clusters for sows, suckling piglets, and weaners. The greatest diversity of antibiotics used was found in weaned piglets, including amoxicillin, lincomycin and enrofloxacin, which were associated with antibiotic resistance genes (mcr1, dfrA1, blaCTX-M). However, direct effects of a particular antibiotic on the associated resistance gene were rare, similarly between antibiotics and virulence genes. However, enrofloxacin and florfenicol favored the occurrence of C. difficile in sows. The Int1 gene was ubiquitously present and could be associated with the prevalence of F4 fimbriae in E. coli, while other E. coli fimbrial genes were not shown to be affected by antibiotics. The vertical transmission of the microbiota from sows to their offspring could be an important factor in the microbial individuality of farms and provide an approach for individual intervention strategies to get piglets through the vulnerable weaning phase in good health. Nevertheless, future studies must include other influencing factors such as the farm environment, detailed health data, and feeding. Extensive databases and integrating artificial intelligence and machine learning can be helpful here. The long-term goal should be indicators for defining a target microbiota that enables the highest possible disease resilience and productivity. It is essential to develop modulation strategies without the use of antibiotics.

Spin systems are known for their purely interacting nature, causing them to lack a canonical perturbative limit, which usually helps to understand the fundamental physics of a system. A method that has proven effective for the description of strongly correlated systems is Functional Renormalization Group (FRG), which can be thought of as an alternative to the path integral formalism as an approach to many-body quantum mechanics. The basic concept is to cure infrared divergences by introducing a cutoff to correlation functions. Continuously lowering the infrared cutoff and hence allowing the system to occupy lower-lying energy states is called the FRG flow and is formulated in terms of differential equations. The solutions to these equations are functions which describe effective interactions between particles and hence contain information about particle correlations and thermodynamic quantities.

Formulating FRG in terms of spins turns out to be a subtle endeavor, though, as the spin algebra relations are relatively complicated compared to, for example, the canonical anti-commutation relations of fermions. In order to use the advantages of FRG for spin systems, one can map spins onto fermionic operators, which leads to pseudo-fermion FRG (PFFRG), developed by Reuther and Wölfle.[198] Alter- natively, one can express spins in terms of Majorana operators, leading to pseudo- Majorana FRG (PMFRG), which has been developed by Niggemann, Sbierski, and Reuther.[163] PMFRG is particularly powerful at finite temperatures, where useful properties of Majorana operators render the method more accurate than PFFRG. In contrast to many other methods, PFFRG and PMFRG are applicable to any spin system, even frustrated ones.

In this work, PMFRG is being generalized and applied to highly frustrated spin systems. In particular, spin representations in terms of Majorana operators for spins with arbitrary large spin magnitude S are being investigated and classified thoroughly, which closes a gap in the literature about the second quantization of spin operators. Moreover, it is presented how to generalize PMFRG for Heisenberg models with full SU(2) symmetry to XXZ models exhibiting only a U(1) symmetry. This opens up a wide range of interesting applications for PMFRG. An example is the XXZ model on the pyrochlore lattice, which is known to exhibit conventional long-range order but also exotic spin liquid ground states in the so-called spin ice phase. PMFRG is applied to this model, and the results are being compared to experiments. Furthermore, the entire phase diagram of the XXZ model is being mapped out. Also, the capability of PMFRG to reproduce low-energy field theory predictions and to determine critical exponents is being tested.

Exultet iam angelica turba coelorum! – Frohlocket, ihr Chöre der Engel! – mit diesen Worten setzt das Osterlob ein; sein Eingangswort Exultet gab den Manuskripten, auf denen das Lob als zentraler Bestandteil der Osterliturgie notiert war, seinen Namen. Die zwischen dem 10. und 14. Jahrhundert vor allem in Apulien und Kampanien entstandenen Exultet-Rollen kombinieren Schrift, Bild und musikalische Notation in einzigartiger Weise.

Das Buch untersucht erstmals materielle und mediale Besonderheiten der Rollen aus interdisziplinärer Perspektive, indem es nicht nur die Objekte selbst, sondern auch ihre Nutzung und Wahrnehmung in der beneventanischen Liturgie in den Blick nimmt. Es wird gezeigt, inwiefern die Rollen visuell, auditiv und olfaktorisch zur Konstitution liturgischer und sozialer Räume beitrugen.

Transform coding methods play a fundamental role in image and video coding technologies like the Versatile Video Coding (VVC) standard. Typically, the employed transforms are linear maps with strong energy compaction capabilities. Therefore, efficient quantization and entropy coding methods can be designed for transmitting and storing the transform coefficients. In recent years, there have been considerable efforts to design coding-efficient transforms from learning-based methods. As for video compression, additional bitrate savings are achieved by optimizing linear block transforms with respect to the different intra prediction modes. In contrast, end-to-end optimized image codecs have been obtained from deep-learning experiments. Learned codecs like the JPEG AI coding standard rely on using non-linear, neural networks as forward and inverse transform. Remarkably, JPEG AI is reported to have superior compression efficiency relative to conventional still image coding. Since the transforms in learned image compression are non-linear, it is not clear if rate-distortion optimization methods designed for linear blocks transforms are well-suited. Thus, this thesis studies the impact of different signal-dependent encoder optimizations on the quantization when a learned image codec is used. As a main result, an algorithm for rate-distortion optimized scalar quantization is developed which achieves bitrate savings between 1 % and 7 %. Furthermore, it has been shown that a rate-constrained vector quantizer improves the coding efficiency on a similar scale. Its design has similarities with the trellis-coded quantization stage in VVC. Thus, since rate-constrained quantization is shown to be effective when applied to non-linear transforms, different non-linear transform coding tools for block-based video compression are developed. These tools employ neural networks which are obtained from a data-driven optimization. The first tool, a non-linear coefficient prediction, uses reconstructed coefficients and the reference samples from the block boundary for predicting low-frequency coefficients. Therefore, only the difference between the predicted value and the original coefficient is quantized and coded. The second tool, a non-linear transform offset, is applied after reconstructing all coefficients and also depends on the reference samples as input. The offset is added before the synthesis transform and has been trained to improve the reconstruction quality. A combination of both methods yields coding gains between 1.0 % and 2.8 % over VVC in All-Intra configuration. Finally, non-linear transforms and intra modes are obtained from an end-to-end training method. The learned transforms do not depend on the reference samples. The training goal is to minimize the expected rate-distortion cost by using an approximation of the transform coefficients’ bitrate. The average All-intra bitrate savings of the learned transforms and intra modes are 0.9 % against VVC.

Autonomous Driving (AD) has advanced significantly in recent years, yet widespread deployment remains limited. One of the most enduring challenges in Autonomous Vehicle (AV) development is planning a safe, comfortable, and efficient motion in complex, real-world environments. This thesis addresses motion planning across three distinct domains: urban shuttles, passenger vehicles, and truck-trailer systems. It contributes practical insights and novel approaches toward scalable autonomous mobility. The first part of this work presents an integrated motion planning framework for the Continental Urban Mobility Experience (CUbE) driverless shuttle. Extensive real-world testing over several years highlights the system’s robustness and underscores the importance of long-term validation in urban settings. Key innovations include a multi-layered planning stack and a data-driven motion forecasting approach that enhances interaction with human traffic participants. The second part investigates the behavior of human drivers in understructured traffic environments. Those are areas that fall between well-defined road systems and fully unstructured spaces. A novel analysis framework is introduced for mining patterns from naturalistic trajectory datasets, enabling AVs to better blend into human traffic and navigate ambiguous scenarios with improved predictability and safety. The final part of the thesis explores Deep Reinforcement Learning (DRL) for planning and controlling complex truck-trailer maneuvers. A DRL-based approach is developed and evaluated in simulated environments, demonstrating the method’s potential to handle the nonlinear dynamics of articulated vehicles. These contributions advance the field of motion planning by combining theoretical insights, system-level integration, and empirical evaluation. They offer pathways for improving AV behavior across diverse platforms and use cases, ultimately supporting the broader adoption of AD technologies.

This dissertation examines Luri oral traditions in Kohgiluyeh and Boyer-Ahmad, focusing on role in shaping historical memory and identity through narratives of Alexander the Great. Using a comparative and interdisciplinary approach, it traces the evolution of Alexander’s legend across Greek, Byzantine, Syriac, and Persian sources, demonstrating that oral traditions are not passive reflections of history but active historiographical forces that preserve, reinterpret, and resist dominant narratives. By analyzing onomastics (naming practices), toponyms (place names), and gendered representations, the study examines how Luri oral traditions function as sites of cultural resistance, challenging Persian nationalist historiography. It explores how Reżā Šāh’s policies marginalized regional oral traditions, promoting the Šāhnāme while sidelining Niẓāmī’s Iskandarnāme, yet Lur oral performers continued to preserve alternative narratives, maintaining distinct historical perspectives. Through ethnographic fieldwork and an analysis of oral storytelling techniques, the study categorizes Luri oral narratives into Mas̩al (proverbs), Muqawwm (chants), Tims̩āl (literary adaptations and genealogies), Matīl (fables), and Iʿtiqāt (beliefs), demonstrating how these traditions actively shape and transmit historical meaning. The findings underscore the urgency of preserving regional oral traditions, as their erosion threatens to erase alternative historiographies and counter-histories essential to understanding Iran’s diverse historical landscape. This study contributes to oral history, historiography, and cultural memory studies, offering insights into how oral traditions function as mechanisms of historical negotiation, cultural resistance, and identity formation beyond Iran.

The Paradoxes of Intimacy explores the formal and narrative possibilities and limitations of writing about intimate human relationships in the second decade of the twenty-first century. In an age marked by globalization, growing social and environmental insecurities, and intensification of capitalist markets (Jeffrey Nealon), the ways we bond with each other have come under increasing scrutiny in scientific research -not least for its links to problems of alienation, solipsism, and social disintegration. Recent sociological research on intimacy has attested to a crisis in the subject describing contemporary forms of close relationships as dramatically cooling (Eva Illouz), unbinding (Zygmunt Baumann), and even eroding (Byung- Chul Han). Meanwhile, the emergence of new forms of intimate bonding, such as same-sex relationships, the #Metoo social movement’s public exposure of private harm, the rise of emotion-aware technologies, and the sentimentalization of capitalist markets have radically transformed the ways intimacies are created and experienced both in actual life and in artful practices. While the subject has been extensively studied in sociological and philosophical disciplines, it is still underexplored in literary scholarship, where intimacy is often used interchangeably with love to refer to either romantic or erotic relationships. Performing a cross-generic analysis of Rachel Cusk’s The Outline Trilogy, Margaret Atwood’s The Heart Goes Last, and Marilynne Robinson’s Lila, The Paradoxes of Intimacy argues that contemporary women’s fiction reimagines the aesthetics of intimacy by moving beyond the exhausted structures of romantic and confessional narratives. Investigating the emergent forms of literary intimacies as they are problematized and re-invented in postmillennial women’s fiction, it critically addresses the question of how literature navigates the liminal spaces of (textual and narrative) encounters and temporalities (of imagined futures and old forms) to dismantle the fictional representation of intimacy as a private matter. While literary intimacies are traditionally predicated on an individualized point of view in which the interiority of the narrator or a character is rendered through introspection, contemporary women’s fiction uses interior monologues and indirect style not to deliver psychological depth, but to reveal the sociality, artificiality, and paradoxicality of connection. In tracing these formal and thematic transformations, the dissertation contends that contemporary women’s fiction does not seek to restore intimacy or propose new modes of authentic bonding. Rather, it reflects and aestheticizes the pervasive disconnection that defines the postmillennial world. Cusk’s autofictional narratives turn intimacy into a performative repetition of stories that reveal the emptiness of self-disclosure; Atwood’s dystopian satire renders love as an apparatus of emotional and economic control; and Robinson’s Lila gestures toward a fragile fleeting alternative that neither resolves nor escapes this crisis. In this sense, The Paradoxes of Intimacy argues that intimacy in contemporary fiction becomes not a site of resolution, but a literary form of crisis and paradox -a mode that exposes the limits of relation, the failure of reciprocity, and the persistent yearning for connection in an age increasingly hostile to the intimate.

Die vorliegende Arbeit befasst sich mit der automatisierten Erkennung und Positionskontrolle von vier spezifischen Arten von Therapiehilfen (Endotrachealtubus bzw. Trachealkanüle, verschiedene Arten von Kathetern, Thoraxdrainagen und Magensonden) in Röntgenthorax Bildern unter Verwendung von tiefen neuronalen Netzwerken. Durch den Einsatz von Segmentationsmodellen und dem Self-Aware-Mechanismus soll die röntgenbasierte Positionskontrolle dieser Therapiehilfen, die in der Intensiv- und Notfallmedizin eine zentrale Rolle spielen, optimiert und automatisiert werden. Dabei werden die Herausforderungen der Positionierung in der klinischen Praxis adressiert, die aufgrund der kritischen Bedeutung der korrekten Positionierung für die Patientensicherheit und Therapieeffizienz von hoher Relevanz sind. Zunächst wurden drei verschiedene methodische Ansätze zur Segmentierung von Therapiehilfen untersucht, die sich in der Komplexität und Spezialisierung der verwendeten Modelle unterscheiden. Dabei zeigte sich, dass der Ansatz mit spezialisierten Modellen für jede Art von Therapiehilfe (Ansatz II) die besten Ergebnisse liefert. Die Evaluierung der Modelle basierte auf den Klassifikationsmetriken AUROC, Sensitivität, Spezifität, Genauigkeit und balancierte Genauigkeit sowie den Segmentierungsmetriken Dice und Dice+, letztere wurde speziell für diese Arbeit entwickelt. Insbesondere die SegResNet-Architektur zeichnete sich durch eine überlegene Segmentationsleistung aus. Ein weiterer Schwerpunkt war die Entwicklung und Evaluierung des Self-Aware-Mechanismus (SA), der in fast allen Szenarien zu einer Verbesserung der Segmentationsleistung führte. Dieser Mechanismus ermöglicht es dem Modell, seine eigenen Vorhersagen zu überprüfen und zu verfeinern, was insbesondere bei der Endpunktverfeinerung und dem Auffüllen oder Verbinden unvollständiger Segmente von Vorteil ist. Die Untersuchung des Einflusses von Trainingsdatenmenge und Bildauflösung auf die Segmentationsleistung ergab, dass sowohl eine größere Datenmenge als auch eine höhere Bildauflösung zu besseren Ergebnissen führen, wobei die Verbesserungen mit zunehmender Datenmenge bzw. Auflösung einem abnehmenden Grenznutzen folgen. Das Ergebnis der Arbeit ist, dass durch den Einsatz von Deep Learning und insbesondere des Self Aware-Mechanismus eine deutliche Leistungssteigerung in der automatisierten Detektion und Positionskontrolle von Therapiehilfen im Röntgenthorax erreicht werden kann. Diese Erkenntnisse bieten wertvolle Ansatzpunkte für die Weiterentwicklung automatisierter Systeme in der medizinischen Bildverarbeitung mit dem Potenzial, die Effizienz und Genauigkeit von automatisierten Kontrollsystemen in der klinischen Praxis signifikant zu verbessern.

Die Habilitationsschrift untersucht neue Aspekte der Einteilung, Diagnostik und Therapie funktioneller und traumatischer Schulterinstabilitäten. Besonders die komplexe funktionelle Schulterinstabilität (FSI) wurde neu charakterisiert und klassifiziert, woraus ein innovativer konservativer Therapieansatz entstand: der Schulter-Schrittmacher. Dieser zeigte in Studien deutliche Vorteile gegenüber der alleinigen Physiotherapie. Zudem wurden moderne Ruhigstellungskonzepte und operative Verfahren verglichen. Die Abduktions-Außenrotationsruhigstellung erwies sich als wirksame Option nach Erstluxation, während operative Stabilisierung niedrigere Rezidivraten bot. Neue operative Techniken, darunter die Bankart-Plus-Methode und der Latarjet-Eingriff nach Voroperationen, wurden hinsichtlich Wirksamkeit und Komplikationen bewertet. Trotz Fortschritten bleiben Fragen zur funktionellen Stabilität, zu Indikationsgrenzen knöcherner Augmentation und zur optimalen Nachbehandlung offen.