The Research Group MITI at TUM University Hospital is seeking a Research Associate / Doctoral Candidate (m/f/d) to support our research on AI-based understanding and representation of surgical procedures in the operating room.
Position overview
Our goal is to further develop our operating room into one of the most highly digitized and data-driven surgical environments in Europe. We are currently building a multimodal data infrastructure that captures data during real-world surgical procedures using multiple cameras, microphones, sensors, and other data sources. Based on this infrastructure, we develop machine learning methods and downstream applications designed to support surgeons and clinical teams in their everyday work. For this position, we are particularly interested in developing structured and real-time representations of ongoing surgical procedures, for example using scene graphs, geometric representations, and multimodal learning approaches. These representations should capture relevant entities, actions, spatial relationships, and temporal developments within the operating room and provide a foundation for intelligent surgical assistance systems.
Tasks and responsibilities
- Participate in innovative research projects at the intersection of medicine, computer science, and artificial intelligence, working closely with experienced clinicians, engineers, and scientists.
- Design and implement novel approaches for operating room and surgical procedure understanding, including scene graphs, graph-based models, and other structured representations.
- Develop machine learning prototypes that integrate multiple data modalities, including 2D/3D vision, language, audio, and sensor data.
- Contribute to the development and expansion of our multimodal operating room infrastructure, with a focus on supporting surgical AI applications and downstream tasks.
- Publish your research in collaboration with colleagues and (inter-) national research partners.
- Pursue a doctoral degree at TUM as part of your research activities.
Qualifications and experience
- Successfully completed university degree (Master's or equivalent) in Computer Science, Mathematics, Robotics, Electrical Engineering, or another relevant technical discipline.
- Strong background in computer vision, machine learning, and deep learning. Experience with geometric deep learning or graph neural networks is highly desirable.
- Strong programming skills, particularly in Python, and experience with relevant frameworks and libraries such as PyTorch, PyTorch Geometric, and OpenCV.
- Practical experience implementing machine learning, computer vision, robotics, or related projects, preferably with the following data types: time-series data, unstructured data, graph-structured data, 2D images/video, or 3D data.
- Ability and willingness to work in an interdisciplinary clinical research environment, including regular work in the operating room.
- Fluent written and spoken English; good German skills are required for working in the clinical environment.
Our offer
- Highly interdisciplinary research environment at the interface of AI, computer vision, medical technology, and surgery.
- Close collaboration with clinicians, computer scientists, engineers, and (inter-) national research partners.
- High degree of scientific freedom and responsibility to develop and pursue own research ideas within research projects.
- Opportunity to publish work at leading international conferences and journals and to participate in scientific conferences.
- Opportunity to pursue a doctoral degree at TUM.
- Employment and remuneration according to TV-L, based on the applicable requirements and qualifications.
Application
Please send your application, including CV, academic transcript of records, and motivation letter to lars.wagner@tum.de.
Contact
Lars Wagner
TUM University Hospital
Department of Surgery – Research Group MITI
Trogerstr. 10
81675 München
Links: Research group MITI, LinkedIn
Die Stelle ist für die Besetzung mit schwerbehinderten Menschen geeignet. Schwerbehinderte Bewerberinnen und Bewerber werden bei ansonsten im wesentlichen gleicher Eignung, Befähigung und fachlicher Leistung bevorzugt eingestellt.