10509 research outputs found
Sort by
Schematic maps for in-car navigation: Schematizing car routes with polygonal landmarks and surrounding street network to improve digital map interaction and spatial knowledge acquisition
No full textDie visuelle Kommunikation aktueller Navigationssysteme ist darauf optimiert, dass der Nutzer passiv einer Anweisung nach der anderen folgt. Sie konzentriert sich auf den Weg und nicht auf die Selbstlokalisierung und Orientierung des Fahrers. In dieser Dissertation wird untersucht, wie schematische Karten Navigationssysteme verbessern können, um den Erwerb von Raumwissen und die Orientierung zu unterstützen. Eine schematische Karte ist eine Art von Kartographie, die sich auf die Topologie der geographischen Elemente konzentriert und nicht auf metrische Genauigkeit. Sie nutzen Generalisierungen, um bestimmte funktionale Aspekte der Karte zu verbessern. Daher haben wir einen neuartigen Algorithmus entwickelt, um Auto-Routenkarten mit Kontextinformationen (Landmarken und das umgebende Straßennetz) zu schematisieren, die darauf abzielen, Orientierungsinformationen hervorzuheben. Die Auswertung der resultierenden Karten hat gezeigt, dass eine solche Visualisierung die Mensch-Computer-Interaktion und den Erwerb von räumlichem Wissen verbessern kann.Current navigation systems' visual communication is optimized for users to follow one instruction at a time passively. It focuses on the way instead of the driver's self-localization and orientation. This dissertation investigates how schematic maps can improve navigation systems to support spatial knowledge acquisition and orientation. A schematic map is a type of cartography that focuses on the geographic elements' topology instead of metric accuracy. They make use of generalizations to improve specific functional aspects of the map. Therefore, we developed a novel algorithm to schematize car route maps with context information (landmarks and the surrounding street network) that is aimed at highlighting orientation information. The resulting maps evaluation demonstrated that such visualization can improve human-computer interaction and spatial knowledge acquisition
Superparamagnetic Nanoparticles with Phosphorescent Complexes as Hybrid Contrast Agents: Integration of MRI and PLIM
No full textTwo different hybrid nanosystems are prepared by loading highly crystalline, monodisperse magnetite nanocubes (MNCs) with phosphorescent Pt(II) complexes (PtCxs). One involves the encapsulation of the hydrophobic PtCx1 within an amphiphilic comb polymer (MNC@poly(maleic anhydride-alt-1-octadecene) [PMAO]–PtCx1), whereas the other involves the direct binding of the hydrophilic PtCx2 to the surface of the MNC mediated by a ligand-exchange procedure (MNC@OH–PtCx2). Both systems are evaluated as potential candidates for multimodal imaging in magnetic resonance imaging (MRI) and photoluminescence lifetime imaging micro(spectro)scopy (PLIM). PLIM measurements on agarose phantoms demonstrate significantly longer excited-state lifetimes compared to the short-lived autofluorescence of biological background. Additionally, both nanosystems perform as effective MRI contrast agents (CAs): the r2* values are 3–4 times higher than for the commercial CA ferucarbotran. MNC@PMAO–PtCx1 particles also cause significant increases in r2. While the ligand exchange procedure efficiently anchors PtCxs to the MNC surface, the polymeric encapsulation ensures higher colloidal stability, contributing to differences in PLIM and MRI outcomes. In these results, the successful integration of two complementary noninvasive imaging modalities within a single nanosystem is confirmed, serving as the impetus for further investigation of such systems as advanced multimodal–multiscale imaging agents with dual orthogonal readouts
Testing the reproducibility of ecological studies on insect behavior in a multi-laboratory setting identifies opportunities for improving experimental rigor
No full textThe reproducibility of studies involving insect species is an underexplored area in the broader discussion about poor reproducibility in science. Our study addresses this gap by conducting a systematic multi-laboratory investigation into the reproducibility of ecological studies on insect behavior. We implemented a 3 × 3 experimental design, incorporating three study sites, and three independent experiments on three insect species from different orders: the turnip sawfly (Athalia rosae, Hymenoptera), the meadow grasshopper (Pseudochorthippus parallelus, Orthoptera), and the red flour beetle (Tribolium castaneum, Coleoptera). Using random-effect meta-analysis, we compared the consistency and accuracy of treatment effects on insect behavioral traits across replicate experiments. We successfully reproduced the overall statistical treatment effect in 83% of the replicate experiments, but overall effect size replication was achieved in only 66% of the replicates. Thus, though demonstrating sufficient reproducibility in some measures, this study also provides the first experimental evidence for cases of poor reproducibility in insect experiments. Our findings further show that reasons causing poor reproducibility established in rodent research also hold for other study organisms and research questions. We believe that a rethinking of current best practices is required to face reproducibility issues in insect studies but also across disciplines. Specifically, we advocate for adopting open research practices and the implementation of methodological strategies that reduce bias and problems arising from over-standardization. With respect to the latter, the introduction of systematic variation through multi-laboratory or heterogenized designs may contribute to improved reproducibility in studies involving any living organisms
Bioinspired crowding directs supramolecular polymerisation
No full textCrowding effects are crucial to maintaining functionality in biological systems, but little is known about their role in analogous artificial counterparts. Within the growing field of supramolecular polymer science, crowding effects have hitherto remained underappreciated. Herein, we show that crowding effects exhibit strong and distinct control over the kinetics, accessible pathways and final outcomes of supramolecular polymerisation processes. In the presence of a pre-formed supramolecular polymer as crowding agent, a model supramolecular polymer dramatically changes its self-assembly behaviour and undergoes a morphological transformation from bundled fibres into flower-like hierarchical assemblies, despite no co-assembly taking place. Notably, this new pathway can only be accessed in crowded environments and when the crowding agent exhibits a one-dimensional morphology. These results allow accessing diverse morphologies and properties in supramolecular polymers and pave the way towards a better understanding of high-precision self-assembly in nature
Challenges in, and recommendations for, hyperspectral imaging in ex vivo malignant glioma biopsy measurements
No full textThe visualization of protoporphyrin IX (PPIX) fluorescence with the help of surgical microscopes during 5-aminolevulinic acid-mediated fluorescence-guided resection (FGR) of gliomas is still limited at the tumor margins. Hyperspectral imaging (HI) detects PPIX more sensitively but is not yet ready for intraoperative use. We illustrate the current status with three experiments and summarize our own experience using HI: (1) assessment of HI analysis algorithm using pig brain tissue, (2) a partially retrospective evaluation of our experience from HI projects, and (3) device comparison of surgical microscopy and HI. In (1), we address the problem that current algorithms for evaluating HI data are based on calibration with liquid phantoms, which have limitations. Their pH is low compared to glioma tissue; they provide only one PPIX photo state and only PPIX as fluorophore. Testing the HI algorithm with brain homogenates, we found proper correction for optical properties but not pH. Considerably more PPIX was measured at pH 9 than at pH 5. In (2), we indicate pitfalls and guide HI application. In (3), we found HI superior to the microscope for biopsy diagnosis (AUC = 0.845 ± 0.024 (cut-off 0.75 µg PPIX/ml) vs. 0.710 ± 0.035). HI thus offers potential for improved FGR
Empirical perspectives on deep learning:from scaling and generalization to new compute paradigms
No full textDie durch neue Rechentechnologien ermöglichte Skalierung künstlicher neuronaler Netze (ANNs) auf nie dagewesene Größen hat nicht nur zu quantitativen Verbesserungen geführt, sondern auch zu qualitativ neuen Phänomenen, die nur bei starker Überparametrisierung zu beobachten sind.
Mit traditionellen theoretischen Ansätzen lassen sich diese großen Modelle nur schwer erfassen.
Um dieses Problem anzugehen, werden in dieser Arbeit empirische Methoden angewandt, um die komplexe, scheinbar chaotische Dynamik großer ANNs zu untersuchen.
Es werden zwei empirische Ansätze vorgestellt, um Generalisierung in ANNs besser zu verstehen und zu verbessern.
Der erste Ansatz untersucht die Breite von Minima in der Loss-Landschaft von ANNs, während der zweite Ansatz die Speicherung einzelner Stichproben durch Vorhersagen zufällig generierter Labels untersucht.
Darüber hinaus haben sich Optische Neuronale Netze (ONNs) als vielversprechende Alternative zu herkömmlichen elektronischen digitalen Rechenarchitekturen herausgestellt.
In dieser Arbeit werden mehrere Strategien für die Implementierung von ONNs vorgeschlagen und es wird erörtert, wie die spezifischen physikalischen Beschränkungen von ONNs aus Sicht der Informatik neue Fragen aufwerfen.Enabled by new computing technologies, the scaling of Artificial Neural Networks (ANNs) to unprecedented sizes has led to more than just quantitative improvements - qualitatively new phenomena have emerged, observable only under heavy overparameterization.
Traditional theoretical approaches struggle to fully capture these large-scale models.
To address this issue, this thesis adopts empirical methods to investigate the complex, seemingly chaotic dynamics of large ANNs.
Two empirical approaches are presented to better understand and improve generalization.
The first approach examines the sharpness of minima in the loss landscape of ANNs, while the second investigates the memorization of individual samples utilizing predictions of randomly generated labels.
Additionally, Optical Neural Networks (ONNs) have emerged as a promising alternative to traditional electronic digital computing architectures.
This thesis proposes several strategies for implementing ONNs and discusses how the specific physical constraints of ONNs raise new questions from a computer science perspective
Functionality and Scar Evaluation of the Donor Site in Extended Radial Forearm Flap Phalloplasty: How Affected Is the Arm in Day-to-Day Life?
No full textBackground: The radial forearm flap remains the gold standard in phalloplasty in gender-affirming surgery due to its versatility and functional outcome, but the significant donor site morbidity and its impact on daily functioning and aesthetic perception remains understudied. This study provides valuable insights into the mid-term functional and aesthetic outcomes of the forearm in transgender individuals following radial forearm flap phalloplasty using widespread instruments for assessment scoring systems and for the evaluation of postoperative wound healing of surgical interventions in general. Methods: Between January 2013 and March 2018, a total of 47 patients underwent radial phalloplasty at AGAPLESION Markus Hospital, and 20 consented to participate in this cross-sectional, retrospective study evaluating functional and aesthetic outcomes post-radial forearm flap phalloplasty using standardized questionnaires (DASH, POSAS, and SBSES). A univariate median regression of each score was performed to determine the associations with selected variables, and correlation analyses between scores was performed using a nonparametric Spearman rank correlation. Results: Among the 20 participants, the median Quick DASH, DASH functionality, DASH sport and music, and DASH work scores indicated minimal to no functional impairment. Scar evaluations using the PSAS, OSAS, and SBSES scales showed overall patient satisfaction with minimal concerns regarding pigmentation and scar texture. Notably, an increase of 0.27 cm2 in wound surface area added one point to the DASH sport and music score (p = 0.037). Statistical analysis also demonstrated a significant correlation between functional and aesthetic assessment scores. Conclusions: The radial forearm phalloplasty donor site, evaluated by the PSAS, OSAS, and SBSES scales, leads to minimal or no functional impairment; however, the worsening of the DASH sport and music score with increasing wound surface reflects a direct relation between wound size and functional impairment
Künstliche Intelligenz als Feedbackassistenz im sozialwissenschaftlichen Unterricht:Eine fachdidaktisch-qualitative Analyse der Potenziale und Grenzen von ChatGPT als lernwirksame Feedbackassistenz für Lernende der Sekundarstufe I am Beispiel der persuasiven Textproduktion im sozialwissenschaftlichen Unterricht
No full textDie vorliegende Bachelorarbeit analysiert die Potenziale und Grenzen von ChatGPT als Feedbackassistenz im sozialwissenschaftlichen Unterricht. Im Fokus steht die Frage, inwiefern ChatGPT die Kriterien eines lernwirksamen Feedbacks zur Förderung persuasiver Argumentationskompetenzen bei Schülerinnen und Schülern der Sekundarstufe I erfüllen kann.
Mithilfe der deduktiven Kategorienanwendung der qualitativen Inhaltsanalyse nach Mayring wurden KI-Feedback-Antworten auf argumentative Texte von Siebtklässlern aus NRW untersucht. Die Ergebnisse zeigen, dass ChatGPT in der Lage ist, formativ-lernwirksames Feedback mit besonderem Fokus auf lernpsychologischen und politikdidaktisch-textsortenspezifischen Aspekten liefern kann, jedoch bei formalen und kognitionspsychologischen Anforderungen an seine Grenzen stößt. Die Arbeit diskutiert anschließend die aus der Analyse resultierenden didaktischen Implikationen sowie die Bedeutung von KI-generiertem Feedback für den sozialwissenschaftlichen Unterricht.This bachelor's thesis analyzes the potential and limitations of ChatGPT as feedback assistance in social science lessons. The focus is on the question of the extent to which ChatGPT can fulfill the criteria of effective feedback for promoting persuasive argumentation skills in lower secondary school students. Using the deductive category application of qualitative content analysis according to Mayring, AI feedback responses to argumentative texts by seventh graders from North Rhine-Westphalia were examined. The results show that ChatGPT is able to provide formative, learning-effective feedback with a particular focus on aspects of learning psychology and political didactics/text type-specific aspects, but reaches its limits when it comes to formal and cognitive psychological requirements. The paper then discusses the didactic implications resulting from the analysis and the significance of AI-generated feedback for social science teaching
Prevalence and quality of temporomandibular disorders, chronic pain and psychological distress in patients with classical and hypermobile Ehlers-Danlos syndrome: an exploratory study
No full textBackground: The Ehlers-Danlos syndromes are a group of clinically and genetically heterogeneous hereditary diseases affecting the connective tissue. They are characterized by hypermobility of the joints, hyperextensible skin and friable tissue. According to current classification, 13 subtypes can be distinguished, of which the hypermobile and the classical subtype are the most prevalent. This study aimed to evaluate patients with classical (cEDS) and hypermobile (hEDS) Ehlers-Danlos syndrome regarding temporomandibular disorder (TMD), chronic pain, and psychological distress. Methods: Support groups from Germany, Austria, and Switzerland were contacted, and social media were used to recruit participants. Free text questions, the German version of the Depression Anxiety and Stress Scale (DASS), and the German version of the Graded Chronic Pain Status (GCPS) were used. Results:
259 participants were included (230 hEDS/29 cEDS). At least 49.2% of the participants had painful or restricted jaw movements, and at least 84.9% had pain in the masticatory muscles, with 46.3% already having a diagnosed TMD. Multivariate analysis showed a significant correlation between TMJ involvement and chronic pain with a 2.5-fold higher risk of chronic pain with a diagnosed TMD. 22.8% of participants had a critical score for depression, 53.3% had a critical score for anxiety, and 34.0% had a critical score for stress. Conclusion: There is a high prevalence of TMD problems and chronic pain in patients with cEDS and hEDS. The lack of knowledge about these problems can create psychological distress. More research is needed to provide adequate treatment for patients with EDS
Propionate metabolism in Desulfurella acetivorans
No full textDesulfurella acetivorans ist ein strikt anaerobes, schwefelreduzierendes Deltaproteobakterium, das heterotroph durch Oxidation von Acetat oder autotroph mit molekularem Wasserstoff wachsen kann. Hier zeigen wir, dass D. acetivorans ein hypothetisches Operon besitzt, das Enzyme des Methylcitrat-Zyklus der Propionat-Oxidation kodiert, und zeigen, dass dieses Bakterium in der Lage ist, auf Propionat zu wachsen. Allerdings konnten die Aktivitäten der Enzyme des Methylcitrat-Zyklus in Extrakten von mit Propionat gewachsenen Zellen nicht nachgewiesen werden, und Experimente mit [U-13C3]-Propionat und eine vergleichende proteomische Analyse von mit Acetat und Propionat gewachsenen Zellen deuteten darauf hin, dass der Methylcitrat-Zyklus während des Propionatwachstums nicht aktiv ist. Stattdessen erfolgt die Propionyl-CoA-Assimilation über die Carboxylierung zu Methylmalonyl-CoA, das weiter zu Succinyl-CoA umgewandelt wird. Letzteres wird in den Citratzyklus geleitet, wo es in Oxalacetat umgewandelt wird und mit Acetyl-CoA (das durch Decarboxylierung eines anderen Oxalacetatmoleküls entsteht) zu Citrat kondensiert, das im Citratyklus oxidiert wird. Diese Ergebnisse verdeutlichen die Unsicherheit genomischer Vorhersagen bei der Analyse mikrobieller Stoffwechselwege und die Notwendigkeit ihrer experimentellen Bestätigung.Desulfurella acetivorans is a strictly anaerobic sulfur-reducing deltaproteobacterium that can grow heterotrophically by oxidation of acetate or autotrophically with molecular hydrogen. Here we show that D. acetivorans possesses a putative operon encoding enzymes of the methylcitrate cycle of propionate oxidation and demonstrate that this bacterium is capable of propionate growth. However, activities of the methylcitrate cycle enzymes could not be detected in extracts of propionate-grown cells, and experiments with [U-13C3]propionate and comparative proteomic analysis of acetate- and propionate-grown cells suggested that the methylcitrate cycle is not active during propionate growth. Instead, propionyl-CoA assimilation proceeds via its carboxylation to methylmalonyl-CoA, which is further converted to succinyl-CoA. The latter is directed to the tricarboxylic acid (TCA) cycle, where it is converted to oxaloacetate and condenses with acetyl-CoA (produced by decarboxylation of another oxaloacetate molecule) to form citrate, which is oxidized in the TCA cycle. These results highlight the uncertainty of genomic predictions in the analysis of microbial metabolic pathways and the need for their experimental confirmation