Investigating applications of machine learning in observing, mapping, and modelling movements of vocal organs for applications within dental healthcar

Currently, the use of dental devices and prosthetics requires individuals to adjust their oral behaviour so that activities such as speech and eating feel as natural as possible. However, such adjustments often end up producing unnatural speech patterns and/or unusual oral movements leading to complications, like an irregular bite.

With the growing interest in healthy ageing and good oral health, new approaches are needed that allow the effective development of customised dental devices and adhesive formulation that will require minimal adjustments, whilst achieving natural oral movement. The primary aim of this research is to develop technologies and approaches that will allow us to realise this goal. To achieve this, detailed models of the mouth and relevant facial/ oral structures will be developed, to help us understand the changes in speech and food processing that occur in a denture-wearing individual. Real time MRI (rtMRI) and Electromagnetic Articulography (EMA) are currently being used as the primary tools for such developments. However, such machinery is expensive, invasive, and labour-intensive; resources that are scarce in dental practices.

This leads to two main research questions. Firstly, is it possible to 3D-model the movements of the tongue during speech and food-processing, by recording only the head and neck with a 360-degree camera? Secondly, how can this technology be applied in visualising the effect of dental prosthetics on one's oral motion?

The research plans to employ Artificial Intelligence and computer vision in creating a system capable of quantifying these changes in speech and food processing, without the need of individual rtMRI and EMA recordings i.e., the possibility of recording the external face and being able to successfully predict the movements of the internal structures responsible for speech (tongue, teeth etc.). The results would then be used to identify common abnormalities occurring from denture use and allow for an appropriate solution. Owing to its multidisciplinary nature, the research would be of great interest to various academic domains. It will entail the development of computer vision tracking algorithms optimised to capture oral behaviour, with the use of a digital camera. These will provide the enhanced capability of capturing movements of the mouth.

It also has significant real-world applications. In the dental field, this may result in a financially and technologically feasible system that can truly allow to produce personalised dental prosthetics, without the associated cost of having to undergo several expensive rtMRI and EMA scans. The technology is not limited to just the face, rather any biological structure that shares a similarity in its internal and external movement patterns could be modelled with such a technology, for example hip joints. It would be possible to diagnose such problems without the need of an MRI. This technology would not eliminate the need of an MRI in any serious scenario but will rather serve as an alternative or supplementary in situations where such scans are currently not feasible. This includes where MRI and EMA machines are not available due to financial and specialist staff requirements.