University of Twente Student Theses


Towards a personalised biomechanical tongue model : sEMG measurements on the tongue and motor unit identification

van Staveren, E.S. (2017) Towards a personalised biomechanical tongue model : sEMG measurements on the tongue and motor unit identification.

Abstract:[Background] The Virtual Therapy Consortium is working on a functional predictive tool to facilitate evidence-based decisions on cancer treatment proposals. For this purpose, a biomechanical model is under development. The biomechanical model will contain high quality 3D animations incorporating patient specific anatomy, physiology, and neuromuscular information. In the case of tongue cancer, it is desired that the model demonstrates patient specific treatment effects on functions as mastication, swallowing, and audible speech. To accomplish these simulations a personalised tongue model is essential. [Objective] The goal of this thesis was to make the first steps towards addition of neuromuscular information to a generic biomechanical tongue model, for personalisation purpose. [Methods] The first step incorporated the development of an electrode setup to measure tongue muscle activation, using the patient-friendly surface electromyography (sEMG). Different electrode setups, using both current materials and new prototypes, were tested for technical and practical suitability. The two best performing grids were applied in an experiment with one healthy subject for tongue sEMG quality evaluation. To distinguish activation of individual tongue muscles, the second step involved identification of motor unit action potential trains (MUAPTs), referred to as decomposition. The recently developed sEMG decomposition approach (KmCKC) of Ning et al. [1] was analysed, optimised and tested on simulated EMGs. This decomposition method was also applied to the experimental tongue sEMG. The third step involved evaluation of the MUAPT propagation patterns for allocation of motor units to individual tongue muscles. [Results] In total, five electrode setups (three current materials and two new prototypes) were evaluated for nine formulated requirements. The existent ECoG grid and prototype Silic-12 grid appeared to be most promising for tongue measurements. During the experiment, the Silic-12 grid showed fewer dislocation artefacts in the tongue sEMG. In the second step, the KmCKC decomposition method showed reasonable results for MUAPT identification in simulated EMGs and provided insight in parameter settings. Unfortunately, its performance in the tongue sEMG could not be verified. The MUAPT propagation patterns over the Silic-12 electrodes allowed some tentative allocations of MU activities to specific superficial intrinsic tongue muscles. [Conclusion] The combination of sEMG measurements on the tongue with the Silic-12 grid and KmCKC decomposition algorithm showed potential for acquisition of neuromuscular information from the superficial intrinsic tongue muscles. Some major improvements should be made in future research before translation to input for the biomechanical tongue model can be initiated.
Item Type:Essay (Master)
Faculty:TNW: Science and Technology
Subject:44 medicine, 50 technical science in general
Programme:Technical Medicine MSc (60033)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page