Deep-learning-based segmentation of the vocal tract and articulators in real-time magnetic resonance images of speech.

Journal: Computer methods and programs in biomedicine
Published Date:

Abstract

BACKGROUND AND OBJECTIVE: Magnetic resonance (MR) imaging is increasingly used in studies of speech as it enables non-invasive visualisation of the vocal tract and articulators, thus providing information about their shape, size, motion and position. Extraction of this information for quantitative analysis is achieved using segmentation. Methods have been developed to segment the vocal tract, however, none of these also fully segment any articulators. The objective of this work was to develop a method to fully segment multiple groups of articulators as well as the vocal tract in two-dimensional MR images of speech, thus overcoming the limitations of existing methods.

Authors

  • Matthieu Ruthven
    Clinical Physics, Barts Health NHS Trust, West Smithfield, London EC1A 7BE, United Kingdom; School of Biomedical Engineering & Imaging Sciences, King's College London, King's Health Partners, St Thomas' Hospital, London SE1 7EH, United Kingdom. Electronic address: matthieuruthven@nhs.net.
  • Marc E Miquel
    Clinical Physics, Barts Health NHS Trust, West Smithfield, London EC1A 7BE, United Kingdom; Centre for Advanced Cardiovascular Imaging, NIHR Barts Biomedical Research Centre, William Harvey Institute, Queen Mary University of London, London EC1M 6BQ, United Kingdom.
  • Andrew P King
    Division of Imaging Sciences and Biomedical Engineering, King's College London, London, United Kingdom. Electronic address: andrew.king@kcl.ac.uk.