Electrode technology for multi-channel EMG recordings from tongue muscles

Tongue is an organ located in the oral cavity that plays a fundamental role in numerous activities such as mastication, swallowing, breathing and speech. It is composed almost entirely of muscle tissue, but despite the relevance of its function, its electro-muscular activity is still poorly analysed and understood. Currently, electrophysiological studies of the tongue are performed with needle electromyography, which represents a limitation both in terms of applicability and capability to extract information at the global muscle level. Due to its functions, requiring a wide range of fine movements, tongue is widely represented at cortical level, therefore, several traumatic events or pathologies involving the central nervous system may impair its function. An appropriate electrophysiological evaluation of this muscle may be relevant in several fields ranging from the diagnosis of neuromuscular disorders to the rehabilitation of specific functions. Tongue fasciculations are one of the signs considered in the diagnosis of Amyotrophic Lateral Sclerosis (ALS). Being able to observe spontaneous activity over a wide tongue area could contribute to an early diagnosis of ALS. Another application concerns the use of tongue activation patterns to assess muscle function or as a biofeedback source in the rehabilitative process of patients with speech defects caused by trauma or strokes. The objective of this thesis was to develop a grid of electrodes for High-Density Surface Electromyography (HDsEMG) detection. The grid was required to cover the entire tongue surface providing, at the same time, the possibility of speech and movement to the subject and it must be able to follow the movements of the tongue, while remaining well attached to its surface. A primary problem consisted in finding a method of adhesion of the electrode grid that was biocompatible, resistant to saliva, while still allowing the acquisition of the sEMG signal. The study was divided in three steps: (i) the characteristics of the acquisition system to be developed were identified and the electrode grid was designed and produced; (ii) an analysis of the impedance associated with the electrode-tongue interface was carried out on 8 subjects during rest; (iii) the grid of electrodes was tested during the pronunciation of a set of individual sounds and letters. The experimental results concerning electrode-tissue impedance showed relatively high (hundreds of kΩ at 50Hz) and highly variable values, within and across subjects. We observed that the presence of glue and the salivation, which tends to make the electrode grid detach, plays a fundamental role in the observed impedance values at the interface. Therefore, in order to obtain a better quality of the acquired signals, it will be necessary to improve the electrodes gluing method. In spite of this, in the cases where the overall contact quality was sufficient, it was possible to acquire sEMG signal and localized the spatial distribution of EMG amplitude associated with specific sounds or letters. From amplitude maps it was possible to observe the activation patterns of the lingual muscles and to identify the size and the location of the active tongue areas. Although preliminary, these results suggest the feasibility and relevance of HDsEMG detection from tongue muscle (at least for short-term recordings). Further studies are required to improve the adhesion of the electrodes to the tongue also for longer term measures.