Personalisation of a biomechanical tongue model for the prediction of treatment outcome.

Among all treatments of cancer, surgery of locally advanced head and neck
cancer has one of the highest risks of loss of vital functions. Speech,
mastication and swallowing are complex functions that are easily affected. A
systematic review on swallowing dysfunction and speech intelligibility after
oral cancer surgery demonstrated that serious swallowing deficiencies were
experienced by over 80% of patients and severe impairment of speech
intelligibility was observed in 20% of patients1. The term *functional
inoperability* is used when unacceptable function loss after surgery is to be
expected2,3. Organ-sparing chemoradiation for advanced oral cancer is used as
an alternative to surgery, but this therapy can also seriously affect vital
functions due to fibrosis and xerostomia4. The choice between these two
treatment modalities is not evidence based. In an international survey among
surgeons and radiotherapists no consensus was reached for the majority of
surgical interventions regarding functional inoperability3. Objective
counselling of patients on the expected functional outcome is therefore
currently impossible. Although severely needed, no standardized evaluation tool
exists to help predict a personalized functional outcome of oral cancer
therapies. In the project *Virtual Therapy*, we are developing such a tool for
oral cancer surgery and radiotherapy. Within the virtual therapy project
(www.virtualtherapy.nl) we are developing a biomechanical model of the oral
cavity to quantitatively predict function loss after treatment (virtual
treatment).

The past couple of years we worked on biomechanical models of the lips and oral
cavity in collaboration with the University of Twente and the University of
British Columbia, which supplied us with a framework (ArtiSynth) to create and
to edit biomechanical models (artisynth.magic.ubc.ca)5. The ArtiSynth framework
is used to develop an advanced surgical simulation tool demonstrating
postoperative movement after a partial glossectomy. click here for a video. A
comparison between the postoperative movement of the model and three patients
showed that the model was indeed able to show qualitative comparable motions6.
This model was also able to simulate impairments as a result of fibrosis which
occurs in both surgery and radiotherapy.

More details can be found in the additional protocol.

The next step is to personalize the model and validate the model using
different modalities. These modalities are grouped into four subprojects.
Finally, the results of the subprojects will be integrated within the
biomechanical model so that a completely personalized virtual partial
glossectomy or radiotherapy treatment can be performed and the posttreatment
impairment can be simulated

Prospective cohort study / Feasibility study

During the three visits the extent of burden will be
- Visit 1: 35 minutes
- Visit 2: 65 minutes (surgery patients) or 20 minutes (radiotherapy patients)
- Visit 3: 20 minutes

None of the procedures in this study are considered to be of any physical harm
to the patient. No invasive measurements are performed.