PhD Projects
The research college Data2Health aims to explore data-driven systems to enable better and earlier medical decisions for optimal therapies and increased patient safety. The necessary collection, storage, transmission and analysis of patient-related data requires trustworthy IT systems. We would like to investigate their conception, design and implementation in this college through a holistic approach with six cooperative doctoral theses in three topics.
PhD Project 1 :
Data Sociality and Data Governance in Healthcare
In this topic area, the organizational and social components of a trustworthy data-driven health ecosystem and the appropriation processes taking place within it are investigated. Innovative technologies such as AI and Big Data, intelligent algorithms and data are understood as infrastructures with "social interaction". Data sociality and data governance structures should contribute to strengthening acceptance and trust as well as to ensuring legal compliance of digital health services.
Aims and research questions:
1. What are the characteristics of data collected, stored, and analyzed in digital health solutions and which theories of action offer explanatory models for underlying data sociality?
2. What data governance structures are required to effectively collaborate in smart healthcare, using trusted AI-based innovative technologies that meet high privacy standards?
PhD student:
NN (Uni Ko)
Supervision:
Prof. Dr. Lutz Thieme (HS Ko)
Prof. Dr. Maria Wimmer (Uni Ko)
Mentor:
Prof. Dr. Bernhard Köppen (Uni Ko)
PhD Project 2 :
Resilience, robustness and trustworthiness of
AI systems in healthcare
AI processes and systems must be secure and resilient for their use in healthcare. Adversarial attacks have occurred in various areas. These attacks can lead to biased, inconsistent, or incorrect results. Therefore, fundamental research is to be conducted on how AI systems in healthcare can be designed securely.
Aims and Research Questions:
1. How can AI procedures in medical use be protected against attacks and manipulation attempts?
2. How can AI-based medical prediction systems be protected?
3. How can users and their domain knowledge be used to improve the resilience of such a prediction system?
PhD student:
Christopher Latz (Uni Ko)
Supervision:
Prof. Dr. Andreas Mauthe (Uni Ko)
Prof. Dr. Maik Kschischo (HS Ko)
Mentor:
Prof. Dr. Jan Jürjens (Uni Ko)
PhD Project 3 :
Clinical decision support to increase patient safety with
AI systems
Unexpected complications associated with serious health damage and even death occur time and again in hospitals. Two examples among many are kidney failure or sepsis. Early signs are sometimes overlooked in everyday hospital life under high work and cost pressure, which can lead to avoidable patient harm. To achieve our goals, we work both with publicly available data and in cooperation with local hospitals of intermediate care levels.
Aims and research questions:
1. Can serious harm events be predicted from hospital patient data using AI methods?
2. How can prior medical knowledge and pre-trained models be integrated?
3. How can such a prediction system be made interpretable, safe, resilient, and robust?
PhD student:
Phillip Wendland (HS Ko)
Supervision:
Prof. Dr. Maik Kschischo (HS Ko)
Prof. Dr. Andreas Mauthe (Uni Ko)
Mentor:
Prof. Dr. Armin Fiedler (HS Ko)
PhD Project 4 :
Sensor technology and learning methods in the early detection of pressure sores (bedsores)
Pressure ulcers occur in people in need of care due to various intrinsic and extrinsic factors. Early detection of risks through processing of camera and sensor data and AI-based warning for caregivers can potentially help prevent such severe damage.
Aims and research questions:
1. can a camera system combined with a new AI prediction model predict grounding events early?
2. How can suitable data be generated for training the AI prediction model?
3. can the prediction system be implemented in a reliable edge computing approach?
PhD student:
NN (HS Ko)
Supervision:
Prof. Dr. Babette Dellen (HS Ko)
Prof. Dr. Uwe Jaekel (HS Ko)
Mentor:
Prof. Dr.-Ing. Dietrich Paulus (Uni Ko)
PhD Project 5 :
Trusted Edge Clouds in Healthcare
The goal of this project is to support tomorrow's hospital, which will rely on edge computing infrastructure to make the most of data-driven technologies for security, efficiency, and latency reasons.
Aims and Research Questions:
1. to leverage the emerging data hubs using edge computing, data will be transferred to local sensors and gateways for the Internet of Things. This can improve the quality of the collected data by performing analysis at the location where the data is collected.
2. in this context, approaches for the secure and trustworthy transfer of sensitive data between the participating companies and institutions support the required distributed data analysis, such as the "International Data Spaces" approach (with instantiation "Medical Data Space") and Gaia-X.
PhD student:
Supervision:
Prof. Dr. Jan Jürjens (Uni Ko)
Prof. Dr. Wolfgang Kiess (HS Ko)
Mentor:
Prof. Dr. Timo Vogt (HS Ko)
PhD Project 6:
Wireless networking in a medical context
The sixth doctoral project is concerned with wireless networking in hospitals. Data derived from the network is to be made usable for medical purposes.
Aims and research questions:
1. How are the data streams of important devices or sensors (e.g., those from PhD project 4) controlled in the multinetwork and connected in parallel as well as coordinated via different wireless technologies to increase their availability?
2. How can medically relevant additional information be derived from the orchestration of wireless technologies to increase trust in the data, for example by assigning vital parameters from mobile sensors to patients?
PhD student:
Supervision:
Prof. Dr. Wolfgang Kiess (HS Ko)
Prof. Dr. Hannes Frey (Uni Ko)
Mentor:
Prof. Dr. Timo Vogt (HS Ko)
