A large portion of patients undergoing tongue tumour resection are bound to return to the OR once more due to tumour regrowth. Healthy tissue margins around the tumour are difficult to define surgically due to the tissue deformation of the tongue. An experimental workflow is set-up to aid the surgeon in reaching negative specimen margins. The futura objective of the experimental workflow is to remove complete tumours in one surgery.
A method to image the specimen intra-operatively is under development. The ex-vivo specimen is intra-operatively 3D scanned with an ultrasound probe. An algorithm automatically segments the tumour in the specimen. Subsequently, the algorithm visualises insufficient specimen margins.
This master assignment aims to develop an imaging method to correlate these virtual margins to the wound bed in order to guide the surgeon in additional resection. Multiple imaging modalities could provide a solution: a pre-operative diagnostic MRI, an in-vivo 3D US scan or an in-vivo 3D light scan. Matching the resected specimen to the wound bed is problematic due to soft tissue deformation. Surgeons report that the tongue tissue tends to shrink after deformation. The wound bed opens up and the specimen shrinks down.
Successful correlation is highly dependen, on a model of this phenomenon. To develop an accurate model, it is empirically deterrnined. After empirica! model development, an experimental implementation phase can be initiated. lt is desired that the empirica! modelling phase and experimental implementation phase utilise the same imaging modality.