Pit Henrich

Artificial Intelligence in Biomedical Engineering, Friedrich-Alexander-University Erlangen-Nürnberg (FAU)

Brief description of the doctoral project

Surgeons typically rely on their anatomical knowledge and preoperative imaging when interacting with organs that can deform significantly during surgery. This reliance on mental modelling can lead to discrepancies between the expected and actual position of anatomical structures. I am exploring the use of neural networks that use sensor data to deform an existing 3D model (e.g. from a CT scan) of the organs to provide interactive 3D information to surgeons or robotic systems. This can help, for example, to track the position of tumours within an organ during an operation. Finally, I am investigating aspects such as model uncertainty in connection with the estimation of deformations.