Magnetic Resonance Imaging (MRI) has been used to provide high-resolution tomographic information, valuable in the study of static vocal tract. However, speech does not present a static behaviour but relies on the continuous and dynamic interaction of the vocal tract articulators and neighbouring tissues. Thus, this could make real-time MRI (rt-MRI) an essential tool to assess speech, with numerous advantages over the current clinical techniques. However, using rt-MRI to image the upper vocal tract remains challenging; the motion of the articulators, including the velum is fast while MRI data acquisition is slow thus inherently limiting temporal resolution. Additionally, an intrinsic loss in SNR, spatial resolution and/or visual image quality is present. The main focus of this thesis is to increase clinical reliability of rt-MRI in speech by investigating novel methodologies for the imaging of velopharyngeal motion. Firstly, commercial rt-MRI protocols at 1.5 T and 3 T were compared, regarding image quality and temporal resolution compromise. Optimal imaging protocols were suggested for both eld strengths. This provided a starting point for future clinical translation and the use of commercial and currently available protocols to image velopharyngeal motion. Furthermore, imaging of velopharyngeal motion was further improved with non-standard acquisition methods, such as non-Cartesian sampling and more advanced reconstruction schemes. An improved imaging protocol for the assessment of velopharyngeal motion was suggested. This was based on a highly accelerated radial trajectory with a novel parallel imaging reconstruction method (radial tt-GRAPPA). The suggested protocol not only allowed for improved image quality and image sharpness,but it was also viable for future clinical translation regarding o offline computation times compared to other reconstruction methodologies also investigated in this thesis. In summary, this thesis added some novel insights into the eld of speech rt-MRI, presenting improved and time effcient imaging protocols, adequate for the assessment of velopharyngeal motion.
Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:766177 |
Date | January 2018 |
Creators | de Freitas, Andreia Calisto |
Publisher | Queen Mary, University of London |
Source Sets | Ethos UK |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Source | http://qmro.qmul.ac.uk/xmlui/handle/123456789/41786 |
Page generated in 0.0016 seconds