ePub :: Startseite

Despite the majority (70%) of female breast cancer cases being diagnosed in women aged 55 years and older and an increasing population of older breast cancer patients due to demographic change, older patients lack adequate representation in cancer research. Thus, this doctoral thesis focused on breast cancer patients aged 50 years and older, aiming to elucidate the role of intratumor heterogeneity, ploidy, and genomic instability and their influences on disease outcome. Therefore, a breast cancer collective of 39 patients with short (median 2.4 years) and long survival (median 19 years) were selected. Multiplex interphase fluorescence in situ hybridization (miFISH) was carried out to analyze copy number alterations (CNAs) of eight breast cancer-associated genes for their potential as biomarkers, as well as assessing genomic instability and tumor heterogeneity. Furthermore, image cytometry was performed to detect ploidy and phylogenetic tree modeling to gain more information about tumor development. Supplementary, targeted next-generation sequencing of 563 breast cancer-associated genes was carried out externally, and the obtained mutation status was statistically analyzed, compared to the miFISH results, and interpreted as part of this thesis. The experimental part of this work revealed several CNAs of breast cancer-specific genes, as well as gene mutations frequently reported in breast carcinomas. The copy number gain of COX2 occurred most frequently (in 72% of the cases), followed by MYC (69%), whereas losses were more common for CDH1 (74%) and TP53 (69%). Comparing aneuploid with diploid tumor samples, significantly higher average signal numbers, CNAs, and instability indices, reflecting the degree of genomic instability, were revealed in the aneuploid tumors. In 16 cases, the signal pattern indicated the formation of an isochromosome 8q and in 14 cases of an isochromosome 17q. Supporting the hypothesis of an isochromosome formation, CNAs of DBC2/MYC and HER2/TP53 significantly co-occurred. Moreover, significant co-occurrence of CNA of HER2/DBC2 was detected, and CNAs for HER2 and PIK3CA mutations and CNAs for CCND1 and PIK3CA mutations were significantly mutually exclusive. Overall, the distribution of gene mutations of the 563-breast cancer-associated genes, as well as the pattern of CNAs in the eight breast cancer-related genes (miFISH) detected in the 39 patients aged 50 years and older, were comparable to results of the age-unbiased TCGA-cohort. Notably, neither the quantity of CNAs, the tumor ploidy, nor the degree of intratumor heterogeneity revealed an association with the survival time, indicating that for patients above the age of 50, these criteria do not seem to have a substantial effect on disease prognosis.

Item

Lipidomic and structural characterization of the cell wall of Streptococcus pneumoniae

(2025-05-13) Alshaar, Belal

Streptococcus pneumoniae (Spn) is a Gram-positive bacterium that naturally colonizes various niches within the human body, but can also lead to infections with varying degrees of symptoms. Despite the availability of vaccines, Spn infections remain a significant health concern, exacerbated by the rising threat of antibiotic resistance. This growing challenge underscores the urgent need to explore new drug targets that are common across different strains. A particularly promising target is the pneumococcal cell wall, which has become a central point of structure-focused research. The cellular membrane, a key component of this wall, plays a crucial role in host-pathogen interactions. Nevertheless, the precise quantities of its components are still not well understood. This thesis presents a first-of-its-kind quantitative analysis of the pneumococcal lipidome, conducted in close collaboration with the group of Prof. Dr. Sven Hammerschmidt from the University of Greifswald. The implementation of a standardized pipeline applicable for in-depth microbial lipidomics on pneumococcal samples is described. Further, I examined how the knockout of selected genes influenced the membrane lipid composition. As a proof of concept, the wild-type serotype 2 strain S. pneumoniae D39 was compared to three mutant strains with single gene knockouts. The targeted mutant strains investigated here are: 1) tacL: TacL is responsible for the attachment of the polymeric teichoic acid chain on the glycolipid anchor, thus forming the lipoteichoic acid (LTA); 2) lgt: Lgt (diacylglyceryl transferase) catalyzes the diacylation of preprolipoproteins via a thioester bond by transferring a diacylglycerol moiety from a phosphatidylglycerol (PG) precursor; 3) cps: enzymes of the cps cluster are responsible for capsule polysaccharide formation. To date, such quantitative lipidome studies have not been performed for Spn. Methodically, shotgun lipidomics was employed and the development of a customized mix of internal standards ensured accurate quantitative analysis. The analysis allowed quantification of more than 100 lipids across the following six classes: diacylglycerol, glucosyldiacylglycerol (GlcDAG), galactosylglucosyldiacylglycerol (GalGlcDAG), cardiolipin, phosphatidylcholine and PG, with GlcDAG and GalGlcDAG being the predominant classes. Gas Chromatography-Mass Spectrometry (GC-MS) and Thin Layer Chromatography in conjunction with high-resolution tandem mass spectrometry were further utilized for the validation of identified lipid species. The lipid composition of tacL and lgt strains showed no significant changes compared to the wild type, while the non-encapsulated strain exhibited significant adaptations. Another aim of this thesis was to optimize a GC-MS method for the accurate quantification of pneumococcal wall glycopolymers, particularly wall teichoic acid and lipoteichoic acid. This is a challenging task for researchers in the field since decades. This quantification technique is anticipated to provide a basis for studying enzymes that affect teichoic acid levels without using radioactive materials in cultures and may also be used for capsule polysaccharide quantification, advancing our understanding of pneumococcal cell wall regulation. Preliminary results from this optimized method have already provided insights into how LytR, which plays an essential role in attaching the wall teichoic acid to the peptidoglycan, influences the teichoic acids ratio. In summary, a robust quantitative method for lipidome analysis has been established, along with a developed method which delivered the first successful proof of concept for teichoic acids quantification in Spn. These approaches enable future investigations into membrane homeostasis and biosynthesis regulation in other mutants or under various growth conditions.

Item

Survival and parasite spread in a spatial host-parasite model with host immunity

(2025) Franck, Sascha Josef

We introduce a stochastic model for the invasion of a parasite population in a spatially structured host population, which includes an individual-based adaptive immune response. We will call this the "Spatial Infection Model with Host Immunity" or SIMI for short. In the SIMI, parasites move as independent simple random walks on a graph until they meet a vertex that is inhabited by a host. With a given probability, the host repels the infection, kills the parasite, and adapts its probability to repel the next infection. After a successful infection attempt, both the host and the attacking parasite die, and only the parasite leaves a random number of offspring. We study the SIMI on the integer line and show that parasites have a positive survival probability if and only if the mean offspring are greater than the mean number of needed infection attempts. Furthermore, we study the speed at which the parasites invade the host population. If the probability of a host after repelling an infection, to also repel the next one does not grow fast enough, then parasites propagate across the host population at a linear speed. However, if that probability grows quickly enough, the propagation speed is polynomial with exponent less than 1. Finally, we investigate the SIMI on higher-dimensional graphs with hosts that are either totally immune and never get infected or get infected in the first attempt. We show that the survival probability undergoes a phase transition in the frequency of totally immune hosts.

Item

Evaluation von ophthalmologischen und systemischen Endpunkten nach Strahlentherapie des posterioren uvealen Melanom

(2025) Erikson, Kristina

Das posteriore uveale Melanom (UM) ist der häufigste intraokuläre primär maligne Tumor, mit 3-5 % Anteil an den malignen Melanomen insgesamt weltweit. Dieses systematische Review konzentriert sich aus klinisch relevanten Gründen auf das Aderhautmelanom und das Ziliarkörpermelanom (zusammengefasst als UM). Bisher gibt es neben altbekannten chirurgischen Methoden, wie der Enukleation, eine Vielzahl an strahlentherapeutischen Möglichkeiten für die Behandlung des UMs. Trotz hinreichender Studienlage ist es bisher allerdings noch nicht gelungen, eindeutig zu identifizieren, welche der Strahlentherapien einen Vorteil in Hinblick auf wichtige klinische Endpunkte, wie die lokale Tumorkontrolle, Metastasierung, Überleben, Visus, Augenerhalt und Strahlennebenwirkungen, hat. Ziel dieser systematischen Übersichtsarbeit und Metaanalyse ist daher der Vergleich der vorhandenen Studienlage, mit Fokus auf die oben genannten klinischen Endpunkte, unter Berücksichtigung der Strahlendosis. Die methodische Vorgehensweise orientiert sich an dem „Cochrane Handbook for Systematic Reviews“ (83). Die Registrierung wurde bei Prospero unter der ID CRD42022311758 durchgeführt. Geeignete Kriterien für den Einschluss relevanter Studien wurden mittels des PICO-Schemas festgelegt. Die PubMed-Datenbank wurde nach Studien durchsucht, die zwischen dem 1. Januar 2000 und dem 31. Dezember 2021 veröffentlicht wurden. Es wurden 20 Studien (18 pro- und retrospektive Kohortenstudien und 2 RCTs) für die weitere Betrachtung identifiziert. Diese flossen in die kritische Bewertung der Studienqualität (Risk of Bias) fand mit den validierten Instrumenten der Cochrane Collaboration ROBINS-I (für nicht-randomisierte Studien) und RoB 2 (für RCTs) ein. Die Metaanalyse für 10 der Studien wurde mit dem Programm R (R-Version 4.1.3, library: meta, procedure: metaprop) (109) durchgeführt. In die Gesamtbewertung fließen die Ergebnisse der Risk of Bias Bewertung, die in der Heterogenitätsanalyse ermittelten Variablen als zusätzliche Einflussgrößen, sowie die in den Forest Plots dargestellten gepoolten Effektschätzer ein. Der kritischste Punkt bei der Erarbeitung unserer Ergebnisse war das Fehlen eines adäquaten Strahlendosisprotokolls mit Angabe der BED und/oder der EQD2. Für die Zukunft sollten Studien prospektiv und randomisiert geplant werden und neben der Angabe der Bestrahlungsdosis pro Augenstruktur auch standardisierte Behandlungsparameter beinhalten, sowie ein einheitliches Nachverfolgungsprotokoll mit eindeutigen, reproduzierbaren Messgrößen.

Item

Invariant integration for prior-knowledge enhanced deep learning architectures

(2025) Rath, Matthias

Incorporating prior knowledge to Deep Neural Networks is a promising approach to improve their sample efficiency by effectively limiting the search space the learning algorithm needs to cover. This reduces the amount of samples a network needs to be trained on to reach a specific performance. Geometrical prior knowledge is the knowledge about input transformations that affect the output in a predictable way, or not at all. It can be built into Deep Neural Networks in a mathematically sound manner by enforcing in- or equivariance. Equivariance is the property of a map to behave predictably under input transformations. Convolutions are an example for a translation-equivariant map, where a translation of the input results in a shifted output. Group-equivariant convolutions are a generalization achieving equivariance towards more general transformation groups such as rotations or flips. Using group-equivariant convolutions within Neural Networks embeds the desired equivariance in addition to translations. Invariance is a closely related concept, where the output of a function does not change when its input is transformed. Invariance is often a desirable property of a feature extractor in the context of classification. While the extracted features need to encode the information required to discriminate between different classes, they should be invariant to intra-class variations, i.e., to transformations that map samples within the same class subspace. In the context of Deep Neural Networks, the required invariant representations can be obtained with mathematical guarantees by applying group-equivariant convolutions followed by globally pooling among the group- and spatial domain. While pooling guarantees invariance, it also discards information and is thus not ideal. In this dissertation, we investigate the transition from equi- to invariance within Deep Neural Networks that leverage geometrical prior knowledge. Therefore, we replace the spatial pooling operation with Invariant Integration, a method that guarantees invariance while adding targeted model capacity rather than destroying information. We first propose an Invariant Integration Layer for rotations based on the group average calculated with monomials. The layer can be readily used within a Neural Network and allows backpropagating through it. The monomial parameters are selected either by iteratively optimizing the least-squared-error of a linear classifier, or based on neural network pruning methods. We then replace the monomials with functions that are more often encountered in the context of Neural Networks such as learnable weighted sums or self-attention. We thereby streamline the training procedure of Neural Networks enhanced with Invariant Integration. Finally, we expand Invariant Integration towards flips and scales, highlighting the universality of our approach. We further propose a multi-stream architecture that is able to leverage invariance to multiple transformations at once. This approach allows us to efficiently combine multiple invariances and select the best-fit invariant solution for the specific problem to solve. The conducted experiments show that applying Invariant Integration in combination with group-equivariant convolutions significantly boosts the sample efficiency of Deep Neural Networks improving the performance when the amount of available training data is limited.

Item

MPC-based vehicle trajectory tracking using machine learning for parameter optimization and fault detection

(2025) Lubiniecki, Toni

This thesis explores advancements in trajectory tracking control and fault detection within automated vehicle systems, focusing on two main areas: developing a learning-based model predictive control algorithm to enhance tracking accuracy and evaluating various neural networks as fault detection systems for trajectory tracking controllers. Both parts are assessed in a high-fidelity simulation environment. The first part presents two adaptive model predictive controllers that use vehicle information, trajectory data, and tracking information to adapt the vehicle model within the model predictive control system, compensating for lost tracking accuracy due to model mismatches. One approach employs a trajectorydynamic lookup table, while the more advanced approach uses Gaussian process regression with clustering. A thorough simulation study on real-world racetracks with varying dynamics demonstrates that the advanced approach effectively manages condition changes, significantly improves tracking performance, handles unknown trajectories with similar improvements, and memorizes adapted behavior through clustering. The second part evaluates the effectiveness of four types of neural networks as fault detection systems. These networks detect changes in the vehicle, environmental shifts, or discrepancies between the applied vehicle model and the real vehicle. Trained a priori through supervised learning, the networks use tracking information, controller outputs, and vehicle data. The evaluation distinguishes between known and unknown fault conditions. The results suggest that neural networks are generally suitable for fault detection systems. Differences in effectiveness among the network types are minor for known fault conditions but more significant for unknown conditions. Integrating adaptive model predictive control and neural network-based fault detection systems shows promise for developing robust and fault-tolerant control systems, enhancing accuracy and maintaining operational integrity in dynamic environments for trajectory tracking.

Item

Hochgradige Aortenklappenstenose mit normaler Ejektionsfraktion

(2025) Kurniadi, Arief

Item

Das Entlassmanagement in der Onkologie und Palliativmedizin

(2025) Kox, Ora-Elena

Item

Rotational atherectomy of heavily calcified coronary lesions

(2025) Mankerious, Nader

Coronary artery calcifications precent in about one fifith of all percutaneous coronary intervention (PCI) procedures [1]. Despite considerable advancements in PCI tools and techniques, coronary calcification remain a significant clinical challenge [2]. Heavily calcified lesions predispose to suboptimal final PCI results [3, 4], underscoring the need for enhanced strategies and tools to adress this challenge. Modified baloons (MB) (cutting and scoring balloons) serve as valuable tools for modifying calcified lesions prior to stent implantation [7]. Nevertheless, these strategies do not achieve optimal plaque modification in over 12% of cases [10, 11]. This highlights the role of rotational atherectomy (RA), which ablates the calcified plaque into small particles (<10 μm), reducing the calcium burden of the lesion and facilitating stent delivery and expansion, potentially improving the long-term performance of drug-eluting stent (DES) [12]. Procedural complications are known to be more frequent among the RA PCIs performed in heavily calcified lesions compared to other PCIs conducted in mild or non-calcified lesions without the use of RA [13]. The SYNTAX score is an anatomically based tool reflecting the cumulative complexity of the coronary tree, taking into account their specific locations and characteristics [15]. We hypothesized that determining the SYNTAX score for the target vessel in the context of RA may serve as a predictive tool for the likelihood of in-hospital adverse outcomes by predicting the technical difficulties faced during the PCI. We have developed the new concept of target vessel SYNTAX score (tvSS) to adress this point. Our analysis aimed to explore the predictors of in-hospital adverse outcomes after RA and to test the target vessel SYNTAX score (tvSS) as a potential causal variable [46]. Patients receiving RA were divided into two groups according to the occurrence of in-hospital adverse outcomes. Median TvSS was significantly higher in patients with vs. without in-hospital adverse outcomes. A tvSS cut-off value of 15 showed 73% sensitivity and 62% specificity for predicting in-hospital adverse outcomes. Moreover, a tvSS ≥15, bailout RA, reduced LVEF, as well as diabetes mellitus, emerged as independent predictors of in-hospital adverse outcomes. After one year follow-up, there was no significant difference between patients with vs. without in-hospital adverse outcomes regarding the rates of composite MACE, all-cause death, MI and TVR. Coronary bifurcation lesions present complex anatomical challenges in the field of interventional cardiology, observed in up to 20% of all PCI cases [18, 19]. They are complex anatomical structures where the main vessel and the side branch (SB) are interpolated within a bifurcation segment [20]. SB compromise is an early described complication during PCI of bifurcation lesions [22]. Main vessel plaques can cause plaque shift leading to side branch compromise [25]. However, the plaque nature including the presence of calcification is also an important predictor of the side branch compromise [26, 27]. The PREPARE-CALC multicenter randomized controlled trial investigated lesion preparation strategies for severely calcified coronary lesions, comparing the use of a MB versus RA, followed by implantation of biodegradable polymer sirolimus eluting stent [11]. We conducted an as-treated subgroup analysis of bifurcation lesions from the PREPARE–CALC trial [47]. We compared outcomes of different modification techniques of severly calcified bifurcation lesion. At the end of the procedure, the SB remained significantly compromised in 15 lesions in MB (32%) and in 5 lesions (7%) in RA group. Fluoroscopy time and procedural duration were significantly higher in RA group. Large coronary dissections were more frequently observed in MB group. In-hospital outcome was similar between both groups. Additionally, we analyzed the evolution of cardiac biomarkers in patients with (n=20) and without (n=84) compromised SB at the end of the procedure until 24 hours. The median value of CK-MB was significantly higher at 16H post-PCI in compromised SB patients with a trend towards higher Troponin T. CTO is observed in approximately one-fifth of all PCI [28]. Heavily calcified CTO lesions present a challenge in CTO PCI and are not uncommon. In fact, moderate to severe calcification characterizes over half of the CTO lesions [29]. The use of RA for the preparation of calcified CTO lesions is increasing [30]. The feasibility and in-hospital outcome of RA in CTOs were reported in some studies [32, 33]. We investigated the in-hospital and long-term outcomes after RA for CTO compared to non-CTO calcified lesions [48]. The angiographic success was less in CTO RA procedures, and the procedural success rate was 80% versus 90.5% in CTO RA versus non-CTO RA procedures. In-hospital MACE was comparable in both groups. However, the incidence of slow flow, coronary perforation and cardiac tamponade were higher in the CTO RA group. We observed a higher 2-year TLF in the CTO group, that was driven by higher cardiac deaths. However, target vessel MI and clinically driven TLR rate were comparable between the study groups. On multivariate regression analysis, the presence of CTO lesion, chronic kidney disease (CKD), periprocedural MI and reduced LV-EF (EF < 50%) were independently associated with the two-year TLF. In the CTO group, 38.7% of patients were treated with elective RA. Compared to elective RA, bail-out procedures took longer time, had higher number of dissections and more frequently required two or more burrs. However, the estimated rate of 2-year TLF was not significantly different between those received elective RA and bailout RA. In conclusion, the use of RA is expanding and reaching high-risk populations. In-hospital adverse outcomes during RA are common in patients with more complex target vessel anatomy as indicated by a higher target vessel SYNTAX score (tvSS ≥ 15). Additionally, bailout RA and left ventricular systolic function emerged as predictors of in-hospital adverse outcomes. In-hospital adverse outcomes were not associated with long-term MACE. In the context of high-risk lesions, comparing a strategy of MB versus RA in severely calcified coronary bifurcation lesions, we observed a significantly higher rate of side branch compromise with an MB-based strategy, which did not translate into worse short-term clinical outcome. Side branch compromise was associated with more extensive periprocedural myocardial injury. Therefore, in calcified bifurcation lesions, an upfront debulking with an RA-based strategy might optimize the result of PCI in the side branch. Furthermore, RA in CTO is feasible with a high success rate. However, the long-term outcome is worse than non-CTO RA. Elective RA in CTO can shorten the procedure time and decrease the incidence of dissection in comparison with bail-out RA. Finally, RA is an indispensable modality in every catheterization laboratory and may be the only solution for lesion preparation in heavily calcified cases, especially in situations where PTCA balloons are unable to cross or dilate the lesion.

Item

Retrospektive, monozentrische Kohortenstudie zur Evaluierung der Wirksamkeit und Sicherheit von SplEntis zur transvaginalen, sakrospinalen FiXierung des primären apikalen PrOlapses (REXO-Studie)

(2025-05-08) Naumann, Lena-Johanna

Item

Wie Ärztinnen und Ärzte ihre Rolle in der Sterbebegleitung und die entsprechenden rechtlich-ethischen Begriffe sehen

(2025) Sommer, Linda

Item

Determinanten der Implementierung von Smartphone-Applikationen für die Langzeitnachsorge nach einer Krebserkrankung im Kindes- bis jungen Erwachsenenalter

(2025) Stamer, Tjorven

Item

Single closed-loop acoustic stimulation during NREM sleep in mice

(2025-05-07) Aksamaz, Sonat

Sleep rhythms can elucidate the neurophysiological processes underlying sleep-dependent memory consolidation, and their neuromodulation. Enhanced slow oscillation (SO) activity, as well as increased phase-coupled spindle and hippocampal sharp wave-ripple (SWR) activity, are associated with improved memory consolidation during sleep across mammalian species. In humans, closed-loop acoustic stimulation (CLAS) has been shown to facilitate the grouping of sleep spindles to SOs using EEG scalp electrodes. In the present study, single closed-loop acoustic stimulation (sCLAS) was implemented in mice, in order to investigate the modulation of SOs, sleep spindles as well as SWRs. Memory retention performance of mice measured in an object place recognition (OPR) task did not differ significantly from the control session under sCLAS, although differential responses were found for Up- versus Down-State Stimulation. Application of sCLAS in the 3h sleep interval did not affect the measured sleep parameters significantly. In all conditions, the acoustic stimuli produced an acute increase in SWR activity followed by a significant suppression of at least 1 second. The only electrophysiological outcome indicating a possible relationship to mnemonic function was the SO-SWR coupling. Further experiments are necessary to investigate the translational relevance of (s)CLAS in mice.

Item

Charakterisierung der Komplement-Zytokin-Interaktion bei Frühgeborenen unter 1500 g Geburtsgewicht mit hohem Sepsisrisiko

(2025) Bienert, Anna-Sophie

Item

Die genetische Landschaft von aggressiven B-Zell-Lymphomen mit MYC- und BCL2- und/oder BCL6-Umlagerungen - Erkenntnisse aus einem präzisionsonkologischen Ansatz und die Rolle des onkologischen Apothekers

(2025-05-08) Riedl, Jörg

Item

Die Entwicklung von Empathie bei Medizinstudierenden im Studienverlauf und assoziierte Faktoren

(2025) Kiehn, Leevke

Empathieförderung steht bisher nicht im Curriculum medizinischer Hochschulen, obwohl die aktuelle Studienlage weltweit zeigt, dass es zu einem Absinken von Empathie im Verlauf des Studiums kommen kann [2]. Die vorliegende Arbeit setzt sich mit der Frage auseinander, ob es auch bei Medizinstudierenden der Universität zu Lübeck zu einer Empathieveränderung im Laufe des Studiums kommt und was es für assoziierte Faktoren gibt. Die zugrundeliegenden Daten wurden im Rahmen der LUST-Studie, einer prospektiv-beobachtenden, monozentrischen und longitudinalen Kohortenstudie, von 2011 bis 2019 erhoben. Bezüglich einer einheitlichen Definition des Begriffes „Empathie“ herrscht international noch keine Einigkeit. Eine häufig verwendete Definition, die auch hier genutzt wurde, ist die von Hojat [36 S. 3]. Er beschreibt Empathie als eine größtenteils kognitive Eigenschaft und grenzt sie bewusst von der Emotionalität ab. Genutzt wurde der, der Definition entsprechenden, validierte Fragenbogen „Jefferson Scale of Empathy – Student Version“ (JSE-S) in deutscher Sprache [36 S. 83]. Neben den demographischen Daten wurden mit Standardinstrumenten zusätzlich Ängstlichkeit und Depressivität (HADS-D), Stress (PMSS-D), Arbeitsbezogene Verhaltens- und Erlebnismuster (AVEM), die Art der Hochschulzulassung, der angestrebte Fachbereich zu Beginn des Studiums und ob die Studierenden eine medizinische Berufsausbildung abgeschlossen haben, erhoben. Die Datenerhebung fand vor Beginn des Studiums und nach zwei, vier und sechs Jahren statt. Die Querschnittstichproben für die verschiedenen Erhebungszeitpunkte lagen zwischen n = 220 und 658 Studierenden und insgesamt 43 Studierende füllten die Baseline- Befragung und alle drei Nachbefragungen aus. Es zeigte sich ein leichter, dennoch statistisch signifikanter Anstieg der Empathiewerte im zeitlichen Verlauf des Studiums. Über fast alle Analysen hinweg waren höhere JSE-S-Summenscores mit dem weiblichen Geschlecht und einem angestrebten „people oriented“ Fachbereich assoziiert. Es konnte ein negativer Zusammenhang zwischen Depressivität und studiumbezogenen Stress mit der Zielvariable JSE-S festgestellt werden. Es gilt zu untersuchen, wieso Studierende mit dem Wunsch ein „technology oriented“ Fachbereich zu ergreifen, geringere Empathiewerte haben als ihre Kommiliton*innen.