Extraction of tongue contour in real-time magnetic resonance imaging sequences

Abstract

Real-time magnetic resonance imaging (rtMRI) is becoming a practical tool in speech production research and language pathology observation. It is still a challenge to extract the tongue contour accurately in rtMRI sequences, since tongue is a soft tissue and often touches other organs such as lips and upper mandible. This paper proposes a novel semiautomatic tongue contour extraction method from rtMRI sequences. The initial boundary image is obtained by combined multi-directional Sobel operators in tongue movement region; then a boundary intensity map is constructed to find the most probable tongue contour points by searching for the optimal boundary route with Viterbi algorithm; finally the tongue contour is obtained using B-Spline approximation. The proposed method could obtain accurate tongue contour from rtMRI sequences, even in the cases that some parts of tongue touch other organs. Experiments demonstrate the robustness of the proposed method.

Segmentation of tongue shapes during vowel production in magnetic resonance images based on statistical modelling

“…A quantitative comparison between the proposed method and other methods in the literature shows that our method is comparable to the one by Zhang et al 13 These authors indicated an average root-mean-square error (RMSE) of 0.74, which is comparable to the mean RMSE of 1.52 obtained using the proposed segmentation method. Peng et al 11 addressed the 2D segmentation of part of the tongue contour, and their results are in accordance with ours.…”

“…The related literature includes studies on intensitybased segmentation [9][10][11][12] and statistical model-based segmentation 4,13 .…”

“…Voice- and speech-related tongue studies are commonly focused on investigating the description of the articulatory dynamics and corresponding acoustic production 9 and on clinical research related to the complete physiology, neurophysiology and structural interplay of the muscular hydrostat complex that is the tongue. The related literature includes studies on intensity-based segmentation 10–13 and statistical model–based segmentation. 5,14…”

Analyzing speech in both time and space: Generalized additive mixed models can uncover systematic patterns of variation in vocal tract shape in real-time MRI

“…However, unlike articulometry data, rt-MRI video frames must first be quantified in some manner before analysis can be carried out. A variety of quantification methods has been proposed (see Ramanarayanan et al, 2018 for a detailed overview), including (but not limited to) region-of-interest analysis (Lammert, Ramanarayanan, Proctor, & Narayanan, 2013;Teixeira et al, 2012;Tilsen et al, 2016), grid-based area or distance functions (Barlaz, Shosted, Fu, & Sutton, 2018;Proctor, Bone, Katsamanis, & Narayanan, 2010;Zhang et al, 2016), image cross-correlation (Lammert, Proctor, & Narayanan, 2010), region-based principal components analysis (Carignan et al, 2019(Carignan et al, , 2015, and automated segmentation of individual speech articulators (Eryildirim, M.-O., & Berger, M.-O., 2011;Labrunie et al, 2018;Silva & Teixeira, 2015).…”

Air-Tissue Boundary Segmentation in Real-Time Magnetic Resonance Imaging Video Using Semantic Segmentation with Fully Convolutional Networks

“…Toutios et al [23] and Sorensen et al [24] used factor analysis technique to estimate the compact outline of the vocal tract. Zhang et al [25] used multi-directional Sobel operators in order to construct boundary intensity map in the rtMRI video frames. Techniques such as [15], [18], [20], [21] are advantageous over the others because of their unsupervised and semi-automatic approach.…”