Publikationen

2024

• Prediction of tumor board procedural recommendations using large language models

  Aubreville, M. ., Ganz, J. ., Ammeling, J. ., Rosbach, E. ., Gehrke, T. ., Scherzad, A. ., … Goncalves, M. . (2024). Prediction of tumor board procedural recommendations using large language models. European Archives of Oto-Rhino-Laryngology. http://doi.org/10.1007/s00405-024-08947-9 (Original work published 2024)

  Abstract

  anzeigen ▸

  Multidisciplinary tumor boards are meetings where a team of medical specialists, including medical oncologists, radiation oncologists, radiologists, surgeons, and pathologists, collaborate to determine the best treatment plan for cancer patients. While decision-making in this context is logistically and cost-intensive, it has a significant positive effect on overall cancer survival.

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Abteilung/Einrichtung

    • FLAIR - Flensburg Artificial Intelligence Research

  • Profilbereich

    • Informatik und Gestaltung

  • Forschungsfeld

    • Digitalisierung im Gesundheitswesen
    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 3: Information und Kommunikation

  Filter Tags

  • 2024
  • Journal Article
  • • Marc Aubreville
    • Jonathan Ganz
    • Jonas Ammeling
    • Emely Rosbach
    • Thomas Gehrke
    • Agmal Scherzad
    • Stephan Hackenberg
    • Miguel Goncalves

• Re-identification from histopathology images

  Ganz, J. ., Ammeling, J. ., Jabari, S. ., Breininger, K. ., & Aubreville, M. . (2024). Re-identification from histopathology images. Medical Image Analysis, 103335. http://doi.org/10.1016/j.media.2024.103335 (Original work published 2024)

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Abteilung/Einrichtung

    • FLAIR - Flensburg Artificial Intelligence Research

  • Forschungsfeld

    • Digitalisierung im Gesundheitswesen
    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 3: Information und Kommunikation

  Filter Tags

  • 2024
  • Journal Article
  • • Jonathan Ganz
    • Jonas Ammeling
    • Samir Jabari
    • Katharina Breininger
    • Marc Aubreville

• QUIC(K) communication for GPU Virtualization in Edge Computing

  Lübben, R. ., Gaikwad, N. B., Stefanović, Čedomir ., & Kosta, S. . (2024). QUIC(K) communication for GPU Virtualization in Edge Computing. In International Conference on Wireless and Mobile Computing, Networking and Communications. Paris: IEEE. (Original work published Oktober 2024)

  • Suche auf Google Scholar

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Lübben, Ralf

  • Abteilung/Einrichtung

    • FLAIR - Flensburg Artificial Intelligence Research

  • Forschungsfeld

    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 4: Wirtschaft

  Filter Tags

  • 2024
  • Conference Paper
  • • Ralf Lübben
    • Nikhil Gaikwad
    • Čedomir Stefanović
    • Sokol Kosta

• Enhancing Drone Operation Efficiency and Operator Experience: Integrating Extended Reality and Adaptive Systems with Situation-Aware Models

  Bendig, H. . (2024). Enhancing Drone Operation Efficiency and Operator Experience: Integrating Extended Reality and Adaptive Systems with Situation-Aware Models. Mensch und Computer 2024. Karlsruhe, Germany: Gesellschaft für Informatik e.V. http://doi.org/10.18420/muc2024-mci-dc-178 (Original work published September 2024)

  Abstract

  anzeigen ▸

  This research project investigates how adaptive Augmented Reality (AR) applications can improve the efficiency of drone operations and the experience of operators. Drones, particularly Unmanned Aerial Vehicles (UAVs), are increasingly employed in search and rescue (SAR) missions due to their rapid deployment capabilities and the ability to access hazardous areas or viewpoints that are otherwise unreachable. Operating drones in these high-demanding scenarios requires operators to possess certain cognitive abilities to manage the UAV effectively. Our research aims to develop a situation-aware interaction model that considers data from the task, the UAVs, and the human operator, offering support during demanding missions or low abilities. This model will be integrated into an AR application designed to adjust to the current situation, thereby enhancing operational outcomes and user experience.

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Bendig, Henner

  • Abteilung/Einrichtung

    • CIVU – Center for Interaction, Visualization and Usability
    • UsabilityLab

  • Profilbereich

    • Informatik und Gestaltung

  • Forschungsfeld

    • Human-Computer-Interaction
    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 3: Information und Kommunikation

  Filter Tags

  • 2024
  • Conference Proceedings
  • • Henner Bendig
  • • augmented reality
    • uav interaction
    • eyetracking
    • biocybernetic control loop
    • Immersive Technologies

• Enhancing Spatial Problem-Solving Through Data-Driven Methods: A Practical Approach Utilizing Immersive Technologies and Multimodal Physical Tracking

  Bendig, H. . (2024). Enhancing Spatial Problem-Solving Through Data-Driven Methods: A Practical Approach Utilizing Immersive Technologies and Multimodal Physical Tracking. In Companion Proceedings of the 29th International Conference on Intelligent User Interfaces (S. 150–152). Greenville, SC, USA: Association for Computing Machinery. http://doi.org/10.1145/3640544.3645244 (Original work published März 2024)

  Abstract

  anzeigen ▸

  This project aims to explore how data-driven methods can be used to enhance humans’ problem-solving skills in spatial tasks and how they can be integrated into immersive tutoring systems. The goal is to provide an enhanced, state-dependent approach that is also applicable outside the laboratory, within real-life scenarios. Digital tutoring and learning systems are typically expert systems and follow a predefined training routine. Ongoing research concentrates on adaptive methods within affective- and cognition-aware systems, yet there is comparatively less investigation into recognizing and adapting problem-solving strategies. Spatial skill have been shown to be especially important for STEM (Science, Technology, Engineering, Mathematics) disciplines in general. Due to their design, immersive systems are particularly suitable for training of spatial skills and are thus frequently utilized for this.

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Bendig, Henner

  • Abteilung/Einrichtung

    • CIVU – Center for Interaction, Visualization and Usability
    • UsabilityLab

  • Profilbereich

    • Informatik und Gestaltung

  • Forschungsfeld

    • Human-Computer-Interaction
    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 3: Information und Kommunikation

  Filter Tags

  • 2024
  • Conference Paper
  • • Henner Bendig
  • • Data-driven Methods
    • Immersive Technologies
    • Spatial Skills
    • biocybernetic control loop

• Hydraulic variable inertia flywheel

  Jauch, C. ., Jost, R. ., & Kloft, P. . (2024). Hydraulic variable inertia flywheel. Applied Energy, 360, 16. http://doi.org/https://doi.org/10.1016/j.apenergy.2024.122830 (Original work published Februar 2024)

  Abstract

  anzeigen ▸

  A novel variable inertia flywheel that uses the mass of a rotating hydraulic fluid is proposed in this paper. In contrast to variable inertia flywheels that use solid masses, this flywheel stands out for its simplicity. In contrast to many other common energy storages, it does not require any environmentally harmful, rare or expensive materials. The basic working principle and the equations that cover the hydraulic behaviour, the pneumatic behaviour and the energy from angular momentum are introduced. These equations are applied to the geometry of the proposed flywheel concept, which allows quantifying the pressures that act on the different mechanical flywheel components. These pressures, together with the centrifugal acceleration from rotation, lead to the loads that have to be withstood by the mechanical components. A simple method for quantifying these loads, and for dimensioning the mechanical components, allows deriving the stationary masses and inertias. A parameter study is conducted, in which different parameters of the flywheel are varied in order to find the maxima in energy density and specific energy. The results of this parameter study reveal that the proposed hydraulic variable inertia flywheel is a very simple and safe energy storage that could provide AC power systems with inertia and control power to support their frequency.

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Jauch, Clemens

  • Forschungsprojekt

    • Inno!Nord: Innovationslabor: Speicher zur Nutzung erneuerbarer Energien im echten Norden

  • Abteilung/Einrichtung

    • Wind Energy Technology Institute

  • Profilbereich

    • Erneuerbare Energien und nachhaltige Entwicklung

  • Forschungsfeld

    • Innovation und Wachstum, Nachhaltigkeit
    • Wind
    • Netzintegration

  • Fachbereich

    • Fachbereich 2: Energy and Life Science

  Filter Tags

  • 2024
  • Journal Article
  • • Prof. Dr. Clemens Jauch
    • Rebecca Jost
    • Peter Kloft
  • • Energy storage
    • Fluid
    • Flywheel
    • Kinetic energy
    • Variable moment of inertia
    • Variable inertia flywheel

• Dominance hierarchy does not influence distances travelled and area utilization in a large group of ponies

  Weidling, G. ., Krieter, J. ., Lübben, R. ., & Czycholl, I. . (2024). Dominance hierarchy does not influence distances travelled and area utilization in a large group of ponies. Applied Animal Behaviour Science, 271. http://doi.org/https://doi.org/10.1016/j.applanim.2024.106178 (Original work published Februar 2024)

  Abstract

  anzeigen ▸

  In order to improve existing horse husbandry practices by providing social contact and exercise opportunities, it is necessary to fully understand the behaviour of horses in relation to group housing. Therefore, this study investigated the influence of dominance hierarchy on the locomotor behaviour and area utilization in a large group of horses in a pasture-based housing system in Northern Germany. The movements of 40 ponies with established group affiliations were observed over approximately one month from late August to early October 2022 using GPS technology. Their dominance hierarchy was determined by testing each possible combination in separate pair-feeding tests. For the determination of area utilization, the available space was divided into a total of 8889 quadrants with 3x3 m each. Analysis of daily distances travelled by the horses revealed an average of 5.4 ± 1.2 km/day, with no significant effect of social rank. However, age had a significant effect, with older horses walking shorter distances. Quadrants used per hour, as measure of area utilization, averaged 37.9 ± 6.4, with no effect of social rank but a significant effect of age observed. Heat maps showed full pasture utilization, with lower ranked horses having less access to resources around the watering trough and a greater use of less attractive grazing areas. The study suggests that in this specific setting, social rank did not significantly affect locomotor behaviour or area use. Although the study did not find a correlation between dominance rank and locomotor behaviour or area use, it highlighted the need for further research in different group sizes and housing systems to optimise group housing practices for horses, especially in settings with limited space and resources.

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Lübben, Ralf

  • Abteilung/Einrichtung

    • FLAIR - Flensburg Artificial Intelligence Research

  • Forschungsfeld

    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 4: Wirtschaft

  Filter Tags

  • 2024
  • Journal Article
  • • Gesa Weidling
    • Joachim Krieter
    • Ralf Lübben
    • Irena Czycholl
  • • Horse
    • Dominance hierarchy
    • GPS
    • Distance
    • Area utilization
    • Group housing

• An Assessment of Costs for System Inertia in Future Power Systems consisting of a high Share of Non-Synchronous Penetration

  Thiesen, H. . (2024). An Assessment of Costs for System Inertia in Future Power Systems consisting of a high Share of Non-Synchronous Penetration. Interdisciplinary Institute for Environmental, Social and Human Sciences Department of Energy and Environmental Management, Europa-Universität Flensburg. Europa-Universität , Flensburg. (Original work published 2024)

  Abstract

  anzeigen ▸

  The need to limit global warming by decarbonising power systems drives the replacement of fossil fuel-fired power plants with renewable energy sources. For the most part, with frequency converter connected wind turbines and photovoltaic systems. As a result, power system inertia decreases. To maintain the controllability of the grid frequency, countermeasures need to be taken, which are associated with costs. This work assesses costs for system inertia in future power systems with a high share of non-synchronous penetration. Publications part of this cumulative thesis are categorised into three phases. The first phase assesses the basics of system inertia and defines fundamental assumptions. In the second phase, single topics such as load inertia, inertia provision via the day-ahead market and the influence of synthetic inertia provided by wind turbines on the power system are assessed. The last phase combines previous findings and analyses system costs of the future German power system due to the provision of inertia. In the Continental European power system, non-synchronous flywheel systems are the least-cost solution to provide necessary synthetic inertia with total annual costs of 167.64e/(kg·m2). Currently, load inertia accounts for 20% of the total system inertia and contribution of different consumer groups varies significantly. Securing sufficient system inertia in the Irish power system via the day-ahead market results in additional operating costs ranging from 1.02e/(kg·m2) to 4.49e/(kg·m2). Wind turbines can provide continuous inertia and are thus able to reduce must-run capacities and resulting CO2 emissions in the Irish power system by 31 %, curtailment by 40% and system costs by 33 %. Results indicate costs in future German power systems for inertia provision in the range from 0.002e/(kg·m2) to 0.61e/(kg·m2). Wind turbines and synchronous condensers equipped with flywheels providing inertia are the most cost-efficient solution to maintain the controllability of the grid frequency in future power systems. Higher CO2 certificate prices need to be taken into account to achieve overall decarbonisation targets. Further, reduction of primary power reserves activation time is crucial.

  • Suche auf Google Scholar

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Thiesen, Henning

  • Abteilung/Einrichtung

    • Wind Energy Technology Institute

  • Profilbereich

    • Erneuerbare Energien und nachhaltige Entwicklung

  • Forschungsfeld

    • Wind
    • Netzintegration

  • Fachbereich

    • Fachbereich 2: Energy and Life Science

  Filter Tags

  • 2024
  • Thesis
  • • Henning Thiesen

• Training for Ultrasound Imaging: An Evaluation of SonoGame

  Gürgan, S. ., Henze, V. ., Gall, C. ., Schoenfisch, B. ., Brucker, S. ., Hahn, M. ., … Teistler, M. . (2024). Training for Ultrasound Imaging: An Evaluation of SonoGame. ACM Symposium on Spatial User Interaction (SUI 2024). Trier, Germany: ACM, New York. http://doi.org/10.1145/3677386.3688880

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Personenseiten

    • Bertel, Sven
    • Teistler, Michael

  • Abteilung/Einrichtung

    • CIVU – Center for Interaction, Visualization and Usability

  • Profilbereich

    • Informatik und Gestaltung

  • Forschungsfeld

    • Human-Computer-Interaction

  • Fachbereich

    • Fachbereich 3: Information und Kommunikation

  Filter Tags

  • 2024
  • Conference Proceedings
  • • Selin Gürgan
    • Victoria Henze
    • Christian Gall
    • Birgitt Schoenfisch
    • Sara Brucker
    • Markus Hahn
    • Sven Bertel
    • Lukas Mayer
    • Matthias Süncksen
    • Michael Teistler

• Breast cancer survival prediction using an automated mitosis detection pipeline

  Stathonikos, N. ., Aubreville, M. ., de Vries, S. ., Wilm, F. ., Bertram, C. A., Veta, M. ., & van Diest, P. J. (2024). Breast cancer survival prediction using an automated mitosis detection pipeline. The Journal of Pathology: Clinical Research, 10, e70008. http://doi.org/https://doi.org/10.1002/2056-4538.70008

  Abstract

  anzeigen ▸

  Abstract Mitotic count (MC) is the most common measure to assess tumor proliferation in breast cancer patients and is highly predictive of patient outcomes. It is, however, subject to inter- and intraobserver variation and reproducibility challenges that may hamper its clinical utility. In past studies, artificial intelligence (AI)-supported MC has been shown to correlate well with traditional MC on glass slides. Considering the potential of AI to improve reproducibility of MC between pathologists, we undertook the next validation step by evaluating the prognostic value of a fully automatic method to detect and count mitoses on whole slide images using a deep learning model. The model was developed in the context of the Mitosis Domain Generalization Challenge 2021 (MIDOG21) grand challenge and was expanded by a novel automatic area selector method to find the optimal mitotic hotspot and calculate the MC per 2 mm2. We employed this method on a breast cancer cohort with long-term follow-up from the University Medical Centre Utrecht (N = 912) and compared predictive values for overall survival of AI-based MC and light-microscopic MC, previously assessed during routine diagnostics. The MIDOG21 model was prognostically comparable to the original MC from the pathology report in uni- and multivariate survival analysis. In conclusion, a fully automated MC AI algorithm was validated in a large cohort of breast cancer with regard to retained prognostic value compared with traditional light-microscopic MC.

  • Suche auf Google Scholar
  • DOI Verweis

  Exportformate

  • BibTeX
  • EndNote X3 XML
  • EndNote 7 XML
  • EndNote tagged
  • RIS

  Weiterführende Links

  anzeigen ▸

  • Abteilung/Einrichtung

    • FLAIR - Flensburg Artificial Intelligence Research

  • Profilbereich

    • Informatik und Gestaltung

  • Forschungsfeld

    • Data Science und Künstliche Intelligenz

  • Fachbereich

    • Fachbereich 3: Information und Kommunikation

  Filter Tags

  • 2024
  • Journal Article
  • • Nikolas Stathonikos
    • Marc Aubreville
    • Sjoerd de Vries
    • Frauke Wilm
    • Christof Bertram
    • Mitko Veta
    • Paul van Diest
  • • artificial intelligence
    • machine learning
    • histopathology
    • prognosis
    • mitosis
    • outcome
    • deep learning