Projects

A. Computer Vision and Medical Imaging 2015-2021:

During my PhD, I was affiliated with the AGAS and VisSim research groups at Koblenz University, Germany, where I obtained my PhD degree in Computer Science. My PhD research was founded upon the Cochlea Model-Based Segmentation (COMBS) project, a funding proposal I co-authored. Funded by the international Cochlear company, this project involved collaboration with the Military Central Hospital (BWZKH) in Koblenz and Ain Shams University Hospital in Egypt. The objective was to assist clinicians in analysing multi-modal 3D medical images of the cochlea for implant candidates. I automated a manual analysis process that previously took experts hours to complete, reducing it to a matter of seconds. This process involved 3D multimodal image fusion, registration, segmentation, and the evaluation of various cochlear internal structures in CBCT, CT, and MRI images. Subsequently, I worked on the PerExt and DynVol projects, focusing on the biomechanical analysis of spine data. I was also actively involved in writing the grant proposals for these projects. I utilised a combination of traditional and AI computer vision methods to segment cervical spine vertebrae in 3D CT and MRI images, as well as to detect muscle and ligament attachment points. This work enabled biomechanical experts to analyse data to improve cervical spine implants and to detect spinal issues early through a simple walking procedure. During this period, I published several scientific papers, delivered presentations at local and international conferences, taught several AI and computer vision courses, and supervised student theses.

B. Data Privacy 2021-2023:

Following my PhD, I joined the Berlin Institute of Health (Charité) as a postdoctoral researcher. As Charité is one of the largest university hospitals in Europe, this role provided a significant opportunity to work on new clinical data challenges. I led our team within the European Joint Programme on Rare Diseases (EJP RD), addressing data privacy for rare-disease medical data. I expanded my expertise in medical data privacy and anonymisation techniques, developing a novel method for generating synthetic data for rare diseases. This work has been published and is gaining interest from the research community. During this time, I also contributed to the development of teaching materials for several courses.

C. Trustworthy AI: 2024-2026

I worked as a project leader representing the RGSE research group in Koblenz on several projects: IH-evrsKI, DATAMITE, and 3DKI. I also contribute to the PrivacyE2E project. My current research focuses on Trustworthy AI topics, e.g., Explainable AI (xAI), AI privacy, and AI fairness

In the IH-evrsKI project, I focus on Explainable AI (xAI) for medical image segmentation and AI fairness metrics, alongside co-organising AI-relevant courses, events, and summer schools. The project aims to enhance students' and society's AI knowledge by providing a general understanding of AI capabilities, limitations, and effective use.

Within DATAMITE, I am developing data valuation metrics and mechanisms for secure data sharing across the dataspace, while actively engaging in project workshops. The project’s objective is to provide a free, open-source framework that empowers companies to value, share, and monetise their data in a legal, ethical, and efficient manner.

Regarding 3DKI, I co-authored our section of the 8-million-euro funding proposal and defined the technical tasks and hardware specifications; this project commenced on 1 November 2025. The goal is to monitor the 3D printing process in real-time and ensure the quality assurance of ceramic objects by analysing Raman, XRT, and point cloud data using AI models.

Finally, for the PrivacyE2E project, I contributed to the design of a conceptual AI privacy model and generated synthetic banking data for the project's case studies. The project aims to provide companies with a framework to ensure legally compliant and ethical AI systems that safeguard user privacy against various attacks.