Precise Markerless Semantic Tracking of Soft Tissues

• Applicants: Professor Dr. Christian Betz; Professor Dr.-Ing. Alexander Schlaefer
• Subject Area: Biomedical Systems Technology ; Otolaryngology, Phoniatrics and Audiology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 528632022

In more complex surgical procedures, it is important to determine the exact position of the target area as well as neighboring structures, for example, to avoid injuring nerves or vessels. For this reason, navigation approaches supporting the surgeons have been established for many types of intervention. In most cases, information on the location of structures is obtained in advance using image data, for example magnetic resonance or computer tomography images. Optical or electromagnetic position tracking systems are typically used to map this information onto patients during surgery. The artificial landmarks used for this purpose can be more easily attached to instruments but are less suitable for determining the location of soft tissues. Localization based on camera images is also often hampered by the lack of features on the tissue surface. In this project, a novel tracking system based on volumetric imaging with fast optical coherence tomography will be developed. Two potential advantages of this approach will be evaluated. On the one hand, the acquisition of signals below the surface can take advantage of natural landmarks even when the surface itself is smooth and lacks significant visible structure. In addition, the high spatial resolution allows comparatively precise localization of the landmarks. On the other hand, the signals can also provide clues about the type of tissue being tracked. Beyond the tissue type, the natural landmarks also have a semantics in terms of their location with respect to the target structures. The use of small image volumes defining natural landmarks offers the advantage of flexible definition of landmarks for tracking. Key research questions of the project concern the selection as well as the fast and precise localization of individual landmarks and the investigation to what extent tissues can be distinguished based on the signal and how this can be considered for the initial placement and when updating landmarks during surgery.

• DFG Programme: Research Grants