- Email: [email protected]
Viscoelastic properties of the tongue and soft palate using MR elastography
Journal of Biomechanics 44 (2011) 450–454
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Viscoelastic properties of the tongue and soft palate using MR elastography S. Cheng a,b, S.C. Gandevia a, M. Green a, R. Sinkus c, L.E. Bilston a,d,n a
Neuroscience Research Australia, University of New South Wales, Hospital Road, Randwick, Sydney 2031, NSW, Australia School of Medical Science, Faculty of Medicine, University of New South Wales, Sydney, Australia Ecole Supe´rieure de Physique et de Chimie Industrielles (ESPCI), Paris, France d Prince of Wales Clinical School, Faculty of Medicine, University of New South Wales, Sydney, Australia b c
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 27 September 2010
Biomechanical properties of the human tongue are needed for finite element models of the upper airway and may be important to elucidate the pathophysiology of obstructive sleep apneoa. Tongue viscoelastic properties have not been characterized previously. Magnetic resonance elastography (MRE) is an emerging imaging technique that can measure the viscoelastic properties of soft tissues in-vivo. In this study, MRE was used to measure the viscoelastic properties of the tongue and soft palate in 7 healthy volunteers during quiet breathing. Results show that the storage shear modulus of the tongue and soft palate is 2.67 7 0.29 and 2.53 7 0.31 kPa (mean 7 SD), respectively. This is the first study to investigate the mechanical properties of the tongue using MRE, and it provides necessary data for future studies of patient groups with altered upper airway function. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Magnetic resonance elastography Tongue mechanical properties Shear modulus Viscoelasticity
1. Introduction The tongue forms a major part of the upper airway and is a complex organ that is critical for multiple physiological tasks such as speech, swallowing and breathing. It is often stated to be a muscular hydrostat, comprised of eight muscles and without any calcified skeletal elements (Gilbert et al., 2007). The eight muscles consist of four extrinsic muscles (genioglossus, hyoglossus, palatoglossus and styloglossus), which attach the tongue to other structures and four intrinsic muscles (superior longitudinal, inferior longitudinal, verticalis and transversus), which lie entirely within the tongue. Each of the physiological tasks is performed by coordinated activation of the different tongue muscles, which results in local stiffening and specific deformation patterns of the lingual musculature. Research interest in the biomechanical modeling of the tongue and the upper airway has grown in recent years, as these models have the potential to elucidate mechanisms of respiratory disorders (eg: obstructive sleep apnea) and used as a simulation tool to predict outcomes from jaw reconstruction and speech therapy (Gerard et al., 2005; Buchaillard et al., 2007; Stavness et al., 2008). Knowledge of tongue mechanical properties is essential for these models. Furthermore, obstructive sleep apnea (OSA) is a heterogeneous disorder, but current methods of diagnosis do not allow differentiation n Corresponding author at: Neuroscience Research Australia, University of New South Wales, Hospital Road, Randwick, Sydney 2031, NSW, Australia. Tel.: + 61 2 9399 1073; fax: +61 2 9399 1027. E-mail addresses: [email protected], [email protected] (L.E. Bilston).
0021-9290/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2010.09.027
between patients whose OSA might be mediated by mechanical factors (Schwab, 2003) or by neural control factors (Strohl, 2003; e.g. high loop gain leading to unstable respiration). For example, despite the association of OSA and high BMI, it is unknown what effect weight gain has on the mechanical properties of the upper airway, specifically whether they alter the stiffness of the tissues, or simply add additional tissue volume (Schwartz et al., 2008). Quadriplegics also develop OSA soon after injury, but they require substantially lower continuous positive airway pressures (CPAP) than able-bodies OSA patients, and MRE might allow us to understand why this is the case by comparing changes in tissue stiffness. Magnetic resonance elastography (MRE) is an emerging imaging technique that measures the viscoelastic properties of soft biological tissues in-vivo (Muthupillai et al., 1995). It propagates mechanical shear waves through the tissue and allows visualization of the waves with an MRI scanner. Tissue properties are extracted by post-processing. MRE has been used to assess tissue properties in several areas of the body such as the breast (Lorenzen et al., 2003), brain (Green et al., 2008), liver (Huwart et al., 2006), heart (Sack et al., 2009), lung and skeletal muscle (Domire et al., 2009; Uffmann et al., 2004; Basford et al., 2002). The purpose of this work was to measure the viscoelastic properties of the human tongue in-vivo using MRE.
2. Methods Seven healthy subjects (6 males and 1 female; age: 25.47 2.6 years (mean7 SD); body mass index: 21.6 7 2.0 kg/m2 (mean 7SD)) were recruited for this study. Subjects lay supine on the scanner bed with a custom-moulded mouth
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454 guard and were requested to keep still and rest their tongue on the hard palate close to the upper teeth during the scans. The subjects were also requested to refrain from swallowing during the scan. This study was approved by the Human Research Ethics Committee of the University of New South Wales and informed written consent was obtained. The study was conducted according to the Declaration of Helsinki. The techniques for the scan sequence and reconstruction methods have been described in detail previously (Sinkus et al., 2000, 2005) and are given here briefly. With the use of motion-sensitive gradients, the three-dimensional displacement fields of synchronized, externally induced mechanical shear waves can be imaged using MRI. As wave propagation (longitudinal and transverse) through a viscoelastic medium is governed by a partial differential wave Eq. (1), which relates the tissue properties to the wave propagation, reconstruction of the viscoelastic properties is possible by analyzing the measured motion field over time -
-
-
-
-
r@2t u ¼ mr2 u þ ðl þ mÞrðr u Þ þ Z@t r2 u þ ðx þ ZÞ@t rðr u Þ
ð1Þ
-
where u (x,t) is the displacement field, r the density of the medium, m the shear modulus, l the second Lame´ coefficient, Z the shear viscosity and x the compressional viscosity. Reconstruction is performed offline using custom software and a short summary of the processing steps for Eq. (1) follows. As soft tissues are almost incompressible, this leads to very small values for the rðr u Þ term and a large magnitude difference between m and l. Eq. (1) is therefore reduced to the following form: -
-
-
r@2t u ¼ mr2 u þ rp þ Z@t r2 u
ð2Þ
-
where p ¼ ðl þ mÞrðr u Þ is the pressure term and is associated with compressional waves. The curl operator is then applied to Eq. (2) to remove the pressure term and this yields a Helmholtz-type equation, which will enable the reconstruction of the shear viscoelastic parameters m and Z at frequency o without the compressional contribution, at the expense of solving third order spatial derivatives. Both the storage modulus, G0 (the elastic component) and loss modulus, G00 (the viscous
MRE coils Mouth guard
451
component) are extracted by numerical solution of the above equation at each voxel from the reconstruction where G0 ¼ m and G00 ¼ oZ. Sagittal scans were performed on a 3T MR scanner (Achieva 1.2; Philips Medical Systems, Best, The Netherlands). Shear waves were propagated through the tongue with a MR-compatible mechanical transducer consisting of two coaxial coils (Green et al., 2008). The transducer (Fig. 1) was mounted on a transmit-receive neurovascular coil and coupled to the maxilla and mandible via a bite bar inserted into an individually moulded polymer mouth guard. The coaxial coils on the transducer were driven by a pulse generator triggered by the MR spectrometer. The associated magnetic field created by the coils is coupled with the magnetic field of the MR scanner to produce a torque and thus oscillation of the coils. The frequency of the oscillation was set at 80 Hz based on preliminary experiments and typical imaging parameters were TR/TE 550/60 ms, scan resolution 64 64 pixels, FOV 200 mm and slice thickness 2 mm. Measurements consisted of 7 slice images leading to a total acquisition time for each dataset of 11 min for each subject. In addition to the MRE dataset, a T2-weighted anatomical scan with identical geometry at higher resolution was collected to identify tongue and soft palate anatomy.
3. Results A movie clip of the shear wave propagation can be viewed in the supplementary file (Movie.mpg). A drawing of the bite bar (yellow) was superimposed onto the movie clip. Supplementary material related to this article can be found online at doi:10.1016/j.jbiomech.2010.09.027. Fig. 2 shows a trace (yellow line) of shear wave penetration in the tongue in the actuation direction of the same subject in the movie clip. Results of the in-vivo viscoelastic properties of the tongue and soft palate for 7 subjects are presented in Table 1 and Fig. 3 shows the data of a typical subject. The mean 7standard deviation values for the tongue and soft palate storage modulus (G0 ) are 2.6770.29 and 2.5370.31 kPa, respectively. The corresponding values for the loss modulus (G00 ) are 0.8570.07 kPa for the tongue and 0.90 70.22 kPa for the soft palate. A paired t-test was used to compare the storage modulus and loss modulus of the tongue and soft palate. Results showed that the shear modulus of the tongue and soft palate are similar and not significantly different (p40.05).
4. Discussion
Fig. 1. The MRE mechanical transducer attached on a neurovascular head coil.
We have estimated the biomechanical properties of the anterior pharynx using MRE at a frequency that is at the lower end of the frequency range in previous MRE studies. It is known that the mechanical properties of soft biological tissues are viscoelastic (Fung, 1993) and their shear modulus increases with test
Displacement (um)
20
10
0
-10
-20 0
10
20
30
40
50
Distance (mm) Fig. 2. The trace (yellow line) of shear wave penetration through the tongue of a typical subject in the actuation direction at an arbitrary time point. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
452
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454
an inverse engineering approach with finite element models of the upper airway. In previous finite element studies, tongue stiffness was estimated by simulation of deformation or collapse at high negative intraluminal pressures (e.g., Xu et al., 2005). The accuracy of the estimates derived from this approach is unknown as the boundary conditions of the human tongue are ambiguous due to its complex coupling with surrounding tissues. Furthermore, the assumed boundary conditions are applied to two-dimensional plane strain models, which simplify the anatomy and mechanics of the upper airway. Table 2 is a summary of the shear modulus of the tongue reported in the literature measured using different methods. There is large variation in tissue properties derived from the finite element approach and between this approach and mechanical test methods. Results from our study correspond well with those of Xu et al. (2005) who derived the mechanical properties of a rodent tongue and soft palate based on a three-dimensional finite element model of the upper airway with realistic boundary conditions obtained using MR tagging (Prince and McVeigh, 1992). The methods used in this study have inherent limitations. Firstly, the mechanical properties of the tongue were reconstructed as an isotropic medium although it is known to be structurally anisotropic in certain regions (Gaige et al., 2007). Tissue stiffness may be anisotropic and further work using an extension of the current technique will allow measurement of stiffness longitudinal and perpendicular to the direction of the muscle fibres (Sinkus et al., 2007). Secondly, due to the difference in oral space between subjects, there are only limited ways that a certain tongue position can be consistently held by all subjects with the mouth guard during the scan. Pilot studies (n¼ 3) were performed to check whether tongue stiffness was affected by its position (Fig. 4 and Table 3) and the results (G0 and G00 ) of the tongue and soft palate stiffness measured with a different tongue position are not significantly different (p40.05). Therefore, it is unlikely that our
frequencies. It is therefore important that readers who wish to compare the stiffness of the tongue relative to other soft biological tissues (measured using MRE) should understand that comparison will only be meaningful if the test frequencies are similar. Tongue stiffness is slightly less than brain (Green et al., 2008) and liver stiffness (Huwart et al., 2006) measured at similar frequencies and analysed using the same reconstruction technique. Measurement of the mechanical properties of the tongue using conventional mechanical methods (eg: tensile test, compression test) is challenging because they require an excised tongue and properties of in-vitro soft tissues are subjected to post-mortem changes in tissue properties (Garo et al., 2007). Such tests also neglect the effects of tonic muscle activation in the tongue (Saboisky et al., 2006). Although in-vivo indentation of the tongue is possible, the muscle has no skeletal support, is sensitive and easily stiffens (due to reflexes) when probed by foreign objects. For these reasons, indirect methods have been used to estimate the biomechanical properties of the tongue and one is Table 1 Summary of the storage modulus (G0 ) and loss modulus (G00 ) of the tongue and soft palate for all the subjects. The measurements are in kPa. Tongue
Soft palate 0
00
Subject
G
G
1 2 3 4 5 6 7 Average SD
2.64 3.09 2.21 2.70 2.94 2.70 2.43 2.67 0.29
0.85 0.77 0.82 0.89 0.97 0.90 0.78 0.85 0.07
Subject
G0
G00
1 2 3 4 5 6 7 Average SD
2.86 2.75 2.15 2.18 2.46 2.89 2.43 2.53 0.31
0.94 0.72 1.17 0.68 0.79 1.23 0.78 0.90 0.22
G’
G” 2.5 2.0 1.5 1.0 0.5 0
Fig. 3. Color maps of storage modulus (G0 ) and loss modulus (G00 ) at an arbitrary time point of a typical subject. Measurements are in kPa.
Table 2 Summary of the mechanical properties (shear modulus) of the tongue and soft palate measured using different techniques. Reference
Tongue (kPa)
Soft Palate (kPa)
Method
Current study Berry et al. (1999) Malhotra et al. (2002) Xu et al. (2005) Liu et al. (2007) Gerard et al. (2005) Birch et al. (2009)
2.67 – 2 1.17 8.33 0.38 –
2.53 0.17–33.5 4 1.67 – – 0.2–0.47
MRE Finite element method Finite element method Finite element method Finite element method Ex-vivo compression Ex-vivo indentation
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454
Position 2
Position 1
453
achieved using MRE with respiratory gating. This study is an important step towards a more complete description of the biomechanics of the tongue and how its mechanical properties change with activation in various physiological tasks, and in the presence of conditions such as OSA.
Conflict of interest statement No conflicts of interest to declare.
Acknowledgements Fig. 4. Different position of the tongue in pilot studies to verify that tongue stiffness is not affected by its position.
Table 3 Storage and loss modulus of the tongue and soft palate measured with a different tongue position. The measurements are in kPa. Position 1 G
0
Position 2 00
G
G
0
This project was funded by a project grant from the National Health and Medical Research Council (NHMRC). We would like to thank Kirsten Moffat and the staff of the Neuroscience Research Australia Clinical Research Imaging Centre for their help. Simon Gandevia and Lynne Bilston are supported by NHMRC research fellowships.
t-test G
References
00
Tongue Average SD
2.75 0.47
0.85 0.10
2.63 0.44
1.02 0.14
p 40.05
Soft palate Average SD
2.45 0.30
0.89 0.24
2.20 0.13
1.09 0.57
p 40.05
results are affected by the tongue position assumed here even though it may not represent the ‘‘usual’’ position of the tongue. Also, although specific instructions to relax the tongue and to breathe quietly during the scan were given, the level of relaxation may have varied between subjects and this may cause inter-subject variation. Three subjects were used to test the intra-subject variation and this was performed with the subjects having different tongue positions. As this study features a new application of MRE, a more thorough intra-subject repeatability test is necessary prior to clinical application of this method. In fact, to the best of our knowledge, intra-subject repeatability test was rarely performed in MRE studies and should be performed to demonstrate the robustness of this technique on new soft tissue. Finally, prior to the study, we were concerned that the vibration stimulus in MRE could evoke muscle activation. In a separate pilot study (Brown et al., 2009), we found that the vibration stimulus had no detectable effect on electromyographic (EMG) activity in genioglossus and thus it is unlikely that the tongue and soft palate stiffness measured were seriously affected by reflex activation due to the vibration stimulus. EMG studies of various tongue muscles (e.g., genioglossus, tensor palatini) during respiration have shown that these muscles are activated to keep the airway patent during inspiration (Sauerland and Harper, 1976). We have recently shown that the mechanical deformation of the tongue occurs during inspiration and is localized to a region just above the geniohyoid and anterior to the epiglottis (Cheng et al., 2008) in young healthy subjects. Therefore, it is probable that stiffness of the tongue is higher in certain tongue regions during inspiration. Such phasic stiffness changes, if they exist, are not accounted for in the current study, where data was acquired over several whole respiratory cycles, thus representing an average stiffness. Further work to measure the active and passive properties of the tongue during respiration is needed to understand the pathophysiology of airway collapse during sleep in OSA. This can be
Basford, J.R., Jenkyn, T.R., An, K.N., Ehman, R.L., Heers, G., Kaufman, K.R., 2002. Evaluation of healthy and diseased muscle with magnetic resonance elastography. Archives of Physical Medicine and Rehabilitation 83 (11), 1530–1536. Berry, D.A., Moon, J.B., Kuehn, D.P., 1999. A finite element model of the soft palate. Cleft-Palate Craniofacial Journal 36 (3), 217–223. Birch, M.J., Srodon, P.D., 2009. Biomechanical properties of the human soft palate. Cleft-Palate Craniofacial Journal 46 (3), 268–274. Brown, E., Cheng, S., McKenzie, D., Gandevia, S., Bilston, L., 2009. Movement and strain of the tongue muscles during awake quiet breathing in obstructive sleep apnoea. Sleep and Biological Rhythms 7, A26. Buchaillard, S., Brix, M., Perrier, P., Payan, Y., 2007. Simulations of the consequences of tongue surgery on tongue mobility: implications for speech production in post-surgery conditions. International Journal of Medical Robotics and Computer Assisted Surgery 3 (3), 252–261. Cheng, S., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2008. Movement of the tongue during normal breathing in awake healthy humans. Journal of Physiology 586, 4283–4294. Domire, Z.J., McCullough, M.B., Chen, Q., An, K.N., 2009. Feasibility of using magnetic resonance elastography to study the effect of aging on shear modulus of skeletal muscle. Journal of Applied Biomechanics 25 (1), 93–97. Fung, Y.C., 1993. Biomechanics: Mechanical Properties of Living Tissues. Springer. Gaige, T.A., Benner, T., Wang, R., Wedeen, V.J., Gilbert, R.J., 2007. Three dimensional myoarchitecture of the human tongue determined in vivo by diffusion tensor imaging with tractography. Journal of Magnetic Resonance Imaging 26 (3), 654–661. Garo, A., Hrapko, M., van Dommelen, J.A., Peters, G.W., 2007. Towards a reliable characterisation of the mechanical behaviour of brain tissue: the effects of postmortem time and sample preparation. Biorheology 44, 51–58. Gerard, J.M., Ohayon, J., Luboz, V., Perrier, P., Payan, Y., 2005. Non-linear elastic properties of the lingual and facial tissues assessed by indentation technique. Application to the biomechanics of speech production. Medical Engineering Physics 27 (10), 884–892. Gilbert, R.J., Napadow, V.J., Gaige, T.A., Wedeen, V.J., 2007. Anatomical basis of lingual hydrostatic deformation. Journal of Experimental Biology 210, 4069–4082. Green, M.A., Bilston, L.E., Sinkus, R., 2008. In vivo brain viscoelastic properties measured by magnetic resonance elastography. NMR in Biomedicine 21 (7), 755–764. Huwart, L., Peeters, F., Sinkus, R., Annet, L., Salameh, N., ter Beek, L.C., Horsmans, Y., Van Beers, B.E., 2006. Liver fibrosis: non-invasive assessment with MR elastography. NMR in Biomedicine 19 (2), 173–179. Liu, Z.S., Luo, X.Y., Lee, H.P., Lu, C., 2007. Snoring source identification and snoring noise prediction. Journal of Biomechanics 40 (4), 861–870. Lorenzen, J., Sinkus, R., Biesterfeldt, M., Adam, G., 2003. Menstrual-cycle dependence of breast parenchyma elasticity: estimation with magnetic resonance elastography of breast tissue during the menstrual cycle. Investigative Radiology 38 (4), 236–240. Malhotra, A., Huang, Y., Fogel, R.B., Pillar, G., Edwards, J.K., Kikinis, R., Loring, S.H., White, D.P., 2002. The male predisposition to pharyngeal collapse: importance of airway length. American Journal of Respiratory and Critical Care Medicine 166 (10), 1388–1395. Muthupillai, R., Lomas, D.J., Rossman, P.J., Greenleaf, J.F., Manduca, A., Ehman, R.L., 1995. Magnetic resonance elastography by direct visualization of propagating acoustic strain waves. Science 269 (5232), 1854–1857.
454
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454
Prince, J.L., McVeigh, E.R., 1992. Motion estimation from tagged MR image sequences. IEEE Transactions on Medical Imaging 11 (2), 238–249. Saboisky, J.P., Butler, J.E., Fogel, R.B., Taylor, J.L., Trinder, J.A., White, D.P., Gandevia, S.C., 2006. Tonic and phasic respiratory drives to human genioglossus motoneurons during breathing. Journal of Neurophysiology 95 (4), 2213–2221. Sack, I., Rump, J., Elgeti, T., Samani, A., Braun, J., 2009. MR elastography of the human heart: noninvasive assessment of myocardial elasticity changes by shear wave amplitude variations. Magnetic Resonance in Medicine 61 (3), 668–677. Sauerland, E.K., Harper, R.M., 1976. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Experimental Neurology 51 (1), 160–170. Schwab, R.J., 2003. Pro: sleep apnea is an anatomic disorder. American Journal of Respiratory and Critical Care Medicine 168, 270–271. Schwartz, A.R., Patil, S.P., Laffan, A.M., Polotsky, V., Schneider, H., Smith, P.L., 2008. Obesity and obstructive sleep apnea: pathogenic mechanisms and therapeutic approaches. The Proceedings of the American Thoracic Society 5, 185–192. Sinkus, R., Lorenzen, J., Schrader, D., Lorenzen, M., Dargatz, M., Holz, D., 2000. High-resolution tensor MR elastography for breast tumour detection. Physics in Medicine and Biology 45 (6), 1649–1664.
Sinkus, R., Siegmann, K., Xydeas, T., Tanter, M., Claussen, C., Fink, M., 2007. MR elastography of breast lesions: understanding the solid/liquid duality can improve the specificity of contrast-enhanced MR mammography. Magnetic Resonance in Medicine 58 (6), 1135–1144. Sinkus, R., Tanter, M., Catheline, S., Lorenzen, J., Kuhl, C., Sondermann, E., Fink, M., 2005. Imaging anisotropic and viscous properties of breast tissue by magnetic resonance-elastography. Magnetic Resonance in Medicine 53 (2), 372–387. Stavness, I., Hannam, A.G., Lloyd, J.E., Fels, S., 2008. Towards predicting biomechanical consequences of jaw reconstruction. Conf Proc IEEE Eng Med Biol Soc, 4567–4570. Strohl, K.P., 2003. American Journal of Respiratory and Critical Care Medicine 168, 271–272. Uffmann, K., Maderwald, S., Ajaj, W., Galban, C.G., Mateiescu, S., Quick, H.H., Ladd, M.E., 2004. In vivo elasticity measurements of extremity skeletal muscle with MR elastography. NMR in Biomedicine 17 (4), 181–190. Xu, C., Brennick, M., Wootton, D., 2005. Image-based three-dimensional finite element modeling approach for upper airway mechanics. Conf Proc IEEE Eng Med Biol Soc 3, 2587–2590.
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Viscoelastic properties of the tongue and soft palate using MR elastography S. Cheng a,b, S.C. Gandevia a, M. Green a, R. Sinkus c, L.E. Bilston a,d,n a
Neuroscience Research Australia, University of New South Wales, Hospital Road, Randwick, Sydney 2031, NSW, Australia School of Medical Science, Faculty of Medicine, University of New South Wales, Sydney, Australia Ecole Supe´rieure de Physique et de Chimie Industrielles (ESPCI), Paris, France d Prince of Wales Clinical School, Faculty of Medicine, University of New South Wales, Sydney, Australia b c
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 27 September 2010
Biomechanical properties of the human tongue are needed for finite element models of the upper airway and may be important to elucidate the pathophysiology of obstructive sleep apneoa. Tongue viscoelastic properties have not been characterized previously. Magnetic resonance elastography (MRE) is an emerging imaging technique that can measure the viscoelastic properties of soft tissues in-vivo. In this study, MRE was used to measure the viscoelastic properties of the tongue and soft palate in 7 healthy volunteers during quiet breathing. Results show that the storage shear modulus of the tongue and soft palate is 2.67 7 0.29 and 2.53 7 0.31 kPa (mean 7 SD), respectively. This is the first study to investigate the mechanical properties of the tongue using MRE, and it provides necessary data for future studies of patient groups with altered upper airway function. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Magnetic resonance elastography Tongue mechanical properties Shear modulus Viscoelasticity
1. Introduction The tongue forms a major part of the upper airway and is a complex organ that is critical for multiple physiological tasks such as speech, swallowing and breathing. It is often stated to be a muscular hydrostat, comprised of eight muscles and without any calcified skeletal elements (Gilbert et al., 2007). The eight muscles consist of four extrinsic muscles (genioglossus, hyoglossus, palatoglossus and styloglossus), which attach the tongue to other structures and four intrinsic muscles (superior longitudinal, inferior longitudinal, verticalis and transversus), which lie entirely within the tongue. Each of the physiological tasks is performed by coordinated activation of the different tongue muscles, which results in local stiffening and specific deformation patterns of the lingual musculature. Research interest in the biomechanical modeling of the tongue and the upper airway has grown in recent years, as these models have the potential to elucidate mechanisms of respiratory disorders (eg: obstructive sleep apnea) and used as a simulation tool to predict outcomes from jaw reconstruction and speech therapy (Gerard et al., 2005; Buchaillard et al., 2007; Stavness et al., 2008). Knowledge of tongue mechanical properties is essential for these models. Furthermore, obstructive sleep apnea (OSA) is a heterogeneous disorder, but current methods of diagnosis do not allow differentiation n Corresponding author at: Neuroscience Research Australia, University of New South Wales, Hospital Road, Randwick, Sydney 2031, NSW, Australia. Tel.: + 61 2 9399 1073; fax: +61 2 9399 1027. E-mail addresses: [email protected], [email protected] (L.E. Bilston).
0021-9290/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2010.09.027
between patients whose OSA might be mediated by mechanical factors (Schwab, 2003) or by neural control factors (Strohl, 2003; e.g. high loop gain leading to unstable respiration). For example, despite the association of OSA and high BMI, it is unknown what effect weight gain has on the mechanical properties of the upper airway, specifically whether they alter the stiffness of the tissues, or simply add additional tissue volume (Schwartz et al., 2008). Quadriplegics also develop OSA soon after injury, but they require substantially lower continuous positive airway pressures (CPAP) than able-bodies OSA patients, and MRE might allow us to understand why this is the case by comparing changes in tissue stiffness. Magnetic resonance elastography (MRE) is an emerging imaging technique that measures the viscoelastic properties of soft biological tissues in-vivo (Muthupillai et al., 1995). It propagates mechanical shear waves through the tissue and allows visualization of the waves with an MRI scanner. Tissue properties are extracted by post-processing. MRE has been used to assess tissue properties in several areas of the body such as the breast (Lorenzen et al., 2003), brain (Green et al., 2008), liver (Huwart et al., 2006), heart (Sack et al., 2009), lung and skeletal muscle (Domire et al., 2009; Uffmann et al., 2004; Basford et al., 2002). The purpose of this work was to measure the viscoelastic properties of the human tongue in-vivo using MRE.
2. Methods Seven healthy subjects (6 males and 1 female; age: 25.47 2.6 years (mean7 SD); body mass index: 21.6 7 2.0 kg/m2 (mean 7SD)) were recruited for this study. Subjects lay supine on the scanner bed with a custom-moulded mouth
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454 guard and were requested to keep still and rest their tongue on the hard palate close to the upper teeth during the scans. The subjects were also requested to refrain from swallowing during the scan. This study was approved by the Human Research Ethics Committee of the University of New South Wales and informed written consent was obtained. The study was conducted according to the Declaration of Helsinki. The techniques for the scan sequence and reconstruction methods have been described in detail previously (Sinkus et al., 2000, 2005) and are given here briefly. With the use of motion-sensitive gradients, the three-dimensional displacement fields of synchronized, externally induced mechanical shear waves can be imaged using MRI. As wave propagation (longitudinal and transverse) through a viscoelastic medium is governed by a partial differential wave Eq. (1), which relates the tissue properties to the wave propagation, reconstruction of the viscoelastic properties is possible by analyzing the measured motion field over time -
-
-
-
-
r@2t u ¼ mr2 u þ ðl þ mÞrðr u Þ þ Z@t r2 u þ ðx þ ZÞ@t rðr u Þ
ð1Þ
-
where u (x,t) is the displacement field, r the density of the medium, m the shear modulus, l the second Lame´ coefficient, Z the shear viscosity and x the compressional viscosity. Reconstruction is performed offline using custom software and a short summary of the processing steps for Eq. (1) follows. As soft tissues are almost incompressible, this leads to very small values for the rðr u Þ term and a large magnitude difference between m and l. Eq. (1) is therefore reduced to the following form: -
-
-
r@2t u ¼ mr2 u þ rp þ Z@t r2 u
ð2Þ
-
where p ¼ ðl þ mÞrðr u Þ is the pressure term and is associated with compressional waves. The curl operator is then applied to Eq. (2) to remove the pressure term and this yields a Helmholtz-type equation, which will enable the reconstruction of the shear viscoelastic parameters m and Z at frequency o without the compressional contribution, at the expense of solving third order spatial derivatives. Both the storage modulus, G0 (the elastic component) and loss modulus, G00 (the viscous
MRE coils Mouth guard
451
component) are extracted by numerical solution of the above equation at each voxel from the reconstruction where G0 ¼ m and G00 ¼ oZ. Sagittal scans were performed on a 3T MR scanner (Achieva 1.2; Philips Medical Systems, Best, The Netherlands). Shear waves were propagated through the tongue with a MR-compatible mechanical transducer consisting of two coaxial coils (Green et al., 2008). The transducer (Fig. 1) was mounted on a transmit-receive neurovascular coil and coupled to the maxilla and mandible via a bite bar inserted into an individually moulded polymer mouth guard. The coaxial coils on the transducer were driven by a pulse generator triggered by the MR spectrometer. The associated magnetic field created by the coils is coupled with the magnetic field of the MR scanner to produce a torque and thus oscillation of the coils. The frequency of the oscillation was set at 80 Hz based on preliminary experiments and typical imaging parameters were TR/TE 550/60 ms, scan resolution 64 64 pixels, FOV 200 mm and slice thickness 2 mm. Measurements consisted of 7 slice images leading to a total acquisition time for each dataset of 11 min for each subject. In addition to the MRE dataset, a T2-weighted anatomical scan with identical geometry at higher resolution was collected to identify tongue and soft palate anatomy.
3. Results A movie clip of the shear wave propagation can be viewed in the supplementary file (Movie.mpg). A drawing of the bite bar (yellow) was superimposed onto the movie clip. Supplementary material related to this article can be found online at doi:10.1016/j.jbiomech.2010.09.027. Fig. 2 shows a trace (yellow line) of shear wave penetration in the tongue in the actuation direction of the same subject in the movie clip. Results of the in-vivo viscoelastic properties of the tongue and soft palate for 7 subjects are presented in Table 1 and Fig. 3 shows the data of a typical subject. The mean 7standard deviation values for the tongue and soft palate storage modulus (G0 ) are 2.6770.29 and 2.5370.31 kPa, respectively. The corresponding values for the loss modulus (G00 ) are 0.8570.07 kPa for the tongue and 0.90 70.22 kPa for the soft palate. A paired t-test was used to compare the storage modulus and loss modulus of the tongue and soft palate. Results showed that the shear modulus of the tongue and soft palate are similar and not significantly different (p40.05).
4. Discussion
Fig. 1. The MRE mechanical transducer attached on a neurovascular head coil.
We have estimated the biomechanical properties of the anterior pharynx using MRE at a frequency that is at the lower end of the frequency range in previous MRE studies. It is known that the mechanical properties of soft biological tissues are viscoelastic (Fung, 1993) and their shear modulus increases with test
Displacement (um)
20
10
0
-10
-20 0
10
20
30
40
50
Distance (mm) Fig. 2. The trace (yellow line) of shear wave penetration through the tongue of a typical subject in the actuation direction at an arbitrary time point. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
452
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454
an inverse engineering approach with finite element models of the upper airway. In previous finite element studies, tongue stiffness was estimated by simulation of deformation or collapse at high negative intraluminal pressures (e.g., Xu et al., 2005). The accuracy of the estimates derived from this approach is unknown as the boundary conditions of the human tongue are ambiguous due to its complex coupling with surrounding tissues. Furthermore, the assumed boundary conditions are applied to two-dimensional plane strain models, which simplify the anatomy and mechanics of the upper airway. Table 2 is a summary of the shear modulus of the tongue reported in the literature measured using different methods. There is large variation in tissue properties derived from the finite element approach and between this approach and mechanical test methods. Results from our study correspond well with those of Xu et al. (2005) who derived the mechanical properties of a rodent tongue and soft palate based on a three-dimensional finite element model of the upper airway with realistic boundary conditions obtained using MR tagging (Prince and McVeigh, 1992). The methods used in this study have inherent limitations. Firstly, the mechanical properties of the tongue were reconstructed as an isotropic medium although it is known to be structurally anisotropic in certain regions (Gaige et al., 2007). Tissue stiffness may be anisotropic and further work using an extension of the current technique will allow measurement of stiffness longitudinal and perpendicular to the direction of the muscle fibres (Sinkus et al., 2007). Secondly, due to the difference in oral space between subjects, there are only limited ways that a certain tongue position can be consistently held by all subjects with the mouth guard during the scan. Pilot studies (n¼ 3) were performed to check whether tongue stiffness was affected by its position (Fig. 4 and Table 3) and the results (G0 and G00 ) of the tongue and soft palate stiffness measured with a different tongue position are not significantly different (p40.05). Therefore, it is unlikely that our
frequencies. It is therefore important that readers who wish to compare the stiffness of the tongue relative to other soft biological tissues (measured using MRE) should understand that comparison will only be meaningful if the test frequencies are similar. Tongue stiffness is slightly less than brain (Green et al., 2008) and liver stiffness (Huwart et al., 2006) measured at similar frequencies and analysed using the same reconstruction technique. Measurement of the mechanical properties of the tongue using conventional mechanical methods (eg: tensile test, compression test) is challenging because they require an excised tongue and properties of in-vitro soft tissues are subjected to post-mortem changes in tissue properties (Garo et al., 2007). Such tests also neglect the effects of tonic muscle activation in the tongue (Saboisky et al., 2006). Although in-vivo indentation of the tongue is possible, the muscle has no skeletal support, is sensitive and easily stiffens (due to reflexes) when probed by foreign objects. For these reasons, indirect methods have been used to estimate the biomechanical properties of the tongue and one is Table 1 Summary of the storage modulus (G0 ) and loss modulus (G00 ) of the tongue and soft palate for all the subjects. The measurements are in kPa. Tongue
Soft palate 0
00
Subject
G
G
1 2 3 4 5 6 7 Average SD
2.64 3.09 2.21 2.70 2.94 2.70 2.43 2.67 0.29
0.85 0.77 0.82 0.89 0.97 0.90 0.78 0.85 0.07
Subject
G0
G00
1 2 3 4 5 6 7 Average SD
2.86 2.75 2.15 2.18 2.46 2.89 2.43 2.53 0.31
0.94 0.72 1.17 0.68 0.79 1.23 0.78 0.90 0.22
G’
G” 2.5 2.0 1.5 1.0 0.5 0
Fig. 3. Color maps of storage modulus (G0 ) and loss modulus (G00 ) at an arbitrary time point of a typical subject. Measurements are in kPa.
Table 2 Summary of the mechanical properties (shear modulus) of the tongue and soft palate measured using different techniques. Reference
Tongue (kPa)
Soft Palate (kPa)
Method
Current study Berry et al. (1999) Malhotra et al. (2002) Xu et al. (2005) Liu et al. (2007) Gerard et al. (2005) Birch et al. (2009)
2.67 – 2 1.17 8.33 0.38 –
2.53 0.17–33.5 4 1.67 – – 0.2–0.47
MRE Finite element method Finite element method Finite element method Finite element method Ex-vivo compression Ex-vivo indentation
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454
Position 2
Position 1
453
achieved using MRE with respiratory gating. This study is an important step towards a more complete description of the biomechanics of the tongue and how its mechanical properties change with activation in various physiological tasks, and in the presence of conditions such as OSA.
Conflict of interest statement No conflicts of interest to declare.
Acknowledgements Fig. 4. Different position of the tongue in pilot studies to verify that tongue stiffness is not affected by its position.
Table 3 Storage and loss modulus of the tongue and soft palate measured with a different tongue position. The measurements are in kPa. Position 1 G
0
Position 2 00
G
G
0
This project was funded by a project grant from the National Health and Medical Research Council (NHMRC). We would like to thank Kirsten Moffat and the staff of the Neuroscience Research Australia Clinical Research Imaging Centre for their help. Simon Gandevia and Lynne Bilston are supported by NHMRC research fellowships.
t-test G
References
00
Tongue Average SD
2.75 0.47
0.85 0.10
2.63 0.44
1.02 0.14
p 40.05
Soft palate Average SD
2.45 0.30
0.89 0.24
2.20 0.13
1.09 0.57
p 40.05
results are affected by the tongue position assumed here even though it may not represent the ‘‘usual’’ position of the tongue. Also, although specific instructions to relax the tongue and to breathe quietly during the scan were given, the level of relaxation may have varied between subjects and this may cause inter-subject variation. Three subjects were used to test the intra-subject variation and this was performed with the subjects having different tongue positions. As this study features a new application of MRE, a more thorough intra-subject repeatability test is necessary prior to clinical application of this method. In fact, to the best of our knowledge, intra-subject repeatability test was rarely performed in MRE studies and should be performed to demonstrate the robustness of this technique on new soft tissue. Finally, prior to the study, we were concerned that the vibration stimulus in MRE could evoke muscle activation. In a separate pilot study (Brown et al., 2009), we found that the vibration stimulus had no detectable effect on electromyographic (EMG) activity in genioglossus and thus it is unlikely that the tongue and soft palate stiffness measured were seriously affected by reflex activation due to the vibration stimulus. EMG studies of various tongue muscles (e.g., genioglossus, tensor palatini) during respiration have shown that these muscles are activated to keep the airway patent during inspiration (Sauerland and Harper, 1976). We have recently shown that the mechanical deformation of the tongue occurs during inspiration and is localized to a region just above the geniohyoid and anterior to the epiglottis (Cheng et al., 2008) in young healthy subjects. Therefore, it is probable that stiffness of the tongue is higher in certain tongue regions during inspiration. Such phasic stiffness changes, if they exist, are not accounted for in the current study, where data was acquired over several whole respiratory cycles, thus representing an average stiffness. Further work to measure the active and passive properties of the tongue during respiration is needed to understand the pathophysiology of airway collapse during sleep in OSA. This can be
Basford, J.R., Jenkyn, T.R., An, K.N., Ehman, R.L., Heers, G., Kaufman, K.R., 2002. Evaluation of healthy and diseased muscle with magnetic resonance elastography. Archives of Physical Medicine and Rehabilitation 83 (11), 1530–1536. Berry, D.A., Moon, J.B., Kuehn, D.P., 1999. A finite element model of the soft palate. Cleft-Palate Craniofacial Journal 36 (3), 217–223. Birch, M.J., Srodon, P.D., 2009. Biomechanical properties of the human soft palate. Cleft-Palate Craniofacial Journal 46 (3), 268–274. Brown, E., Cheng, S., McKenzie, D., Gandevia, S., Bilston, L., 2009. Movement and strain of the tongue muscles during awake quiet breathing in obstructive sleep apnoea. Sleep and Biological Rhythms 7, A26. Buchaillard, S., Brix, M., Perrier, P., Payan, Y., 2007. Simulations of the consequences of tongue surgery on tongue mobility: implications for speech production in post-surgery conditions. International Journal of Medical Robotics and Computer Assisted Surgery 3 (3), 252–261. Cheng, S., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2008. Movement of the tongue during normal breathing in awake healthy humans. Journal of Physiology 586, 4283–4294. Domire, Z.J., McCullough, M.B., Chen, Q., An, K.N., 2009. Feasibility of using magnetic resonance elastography to study the effect of aging on shear modulus of skeletal muscle. Journal of Applied Biomechanics 25 (1), 93–97. Fung, Y.C., 1993. Biomechanics: Mechanical Properties of Living Tissues. Springer. Gaige, T.A., Benner, T., Wang, R., Wedeen, V.J., Gilbert, R.J., 2007. Three dimensional myoarchitecture of the human tongue determined in vivo by diffusion tensor imaging with tractography. Journal of Magnetic Resonance Imaging 26 (3), 654–661. Garo, A., Hrapko, M., van Dommelen, J.A., Peters, G.W., 2007. Towards a reliable characterisation of the mechanical behaviour of brain tissue: the effects of postmortem time and sample preparation. Biorheology 44, 51–58. Gerard, J.M., Ohayon, J., Luboz, V., Perrier, P., Payan, Y., 2005. Non-linear elastic properties of the lingual and facial tissues assessed by indentation technique. Application to the biomechanics of speech production. Medical Engineering Physics 27 (10), 884–892. Gilbert, R.J., Napadow, V.J., Gaige, T.A., Wedeen, V.J., 2007. Anatomical basis of lingual hydrostatic deformation. Journal of Experimental Biology 210, 4069–4082. Green, M.A., Bilston, L.E., Sinkus, R., 2008. In vivo brain viscoelastic properties measured by magnetic resonance elastography. NMR in Biomedicine 21 (7), 755–764. Huwart, L., Peeters, F., Sinkus, R., Annet, L., Salameh, N., ter Beek, L.C., Horsmans, Y., Van Beers, B.E., 2006. Liver fibrosis: non-invasive assessment with MR elastography. NMR in Biomedicine 19 (2), 173–179. Liu, Z.S., Luo, X.Y., Lee, H.P., Lu, C., 2007. Snoring source identification and snoring noise prediction. Journal of Biomechanics 40 (4), 861–870. Lorenzen, J., Sinkus, R., Biesterfeldt, M., Adam, G., 2003. Menstrual-cycle dependence of breast parenchyma elasticity: estimation with magnetic resonance elastography of breast tissue during the menstrual cycle. Investigative Radiology 38 (4), 236–240. Malhotra, A., Huang, Y., Fogel, R.B., Pillar, G., Edwards, J.K., Kikinis, R., Loring, S.H., White, D.P., 2002. The male predisposition to pharyngeal collapse: importance of airway length. American Journal of Respiratory and Critical Care Medicine 166 (10), 1388–1395. Muthupillai, R., Lomas, D.J., Rossman, P.J., Greenleaf, J.F., Manduca, A., Ehman, R.L., 1995. Magnetic resonance elastography by direct visualization of propagating acoustic strain waves. Science 269 (5232), 1854–1857.
454
S. Cheng et al. / Journal of Biomechanics 44 (2011) 450–454
Prince, J.L., McVeigh, E.R., 1992. Motion estimation from tagged MR image sequences. IEEE Transactions on Medical Imaging 11 (2), 238–249. Saboisky, J.P., Butler, J.E., Fogel, R.B., Taylor, J.L., Trinder, J.A., White, D.P., Gandevia, S.C., 2006. Tonic and phasic respiratory drives to human genioglossus motoneurons during breathing. Journal of Neurophysiology 95 (4), 2213–2221. Sack, I., Rump, J., Elgeti, T., Samani, A., Braun, J., 2009. MR elastography of the human heart: noninvasive assessment of myocardial elasticity changes by shear wave amplitude variations. Magnetic Resonance in Medicine 61 (3), 668–677. Sauerland, E.K., Harper, R.M., 1976. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Experimental Neurology 51 (1), 160–170. Schwab, R.J., 2003. Pro: sleep apnea is an anatomic disorder. American Journal of Respiratory and Critical Care Medicine 168, 270–271. Schwartz, A.R., Patil, S.P., Laffan, A.M., Polotsky, V., Schneider, H., Smith, P.L., 2008. Obesity and obstructive sleep apnea: pathogenic mechanisms and therapeutic approaches. The Proceedings of the American Thoracic Society 5, 185–192. Sinkus, R., Lorenzen, J., Schrader, D., Lorenzen, M., Dargatz, M., Holz, D., 2000. High-resolution tensor MR elastography for breast tumour detection. Physics in Medicine and Biology 45 (6), 1649–1664.
Sinkus, R., Siegmann, K., Xydeas, T., Tanter, M., Claussen, C., Fink, M., 2007. MR elastography of breast lesions: understanding the solid/liquid duality can improve the specificity of contrast-enhanced MR mammography. Magnetic Resonance in Medicine 58 (6), 1135–1144. Sinkus, R., Tanter, M., Catheline, S., Lorenzen, J., Kuhl, C., Sondermann, E., Fink, M., 2005. Imaging anisotropic and viscous properties of breast tissue by magnetic resonance-elastography. Magnetic Resonance in Medicine 53 (2), 372–387. Stavness, I., Hannam, A.G., Lloyd, J.E., Fels, S., 2008. Towards predicting biomechanical consequences of jaw reconstruction. Conf Proc IEEE Eng Med Biol Soc, 4567–4570. Strohl, K.P., 2003. American Journal of Respiratory and Critical Care Medicine 168, 271–272. Uffmann, K., Maderwald, S., Ajaj, W., Galban, C.G., Mateiescu, S., Quick, H.H., Ladd, M.E., 2004. In vivo elasticity measurements of extremity skeletal muscle with MR elastography. NMR in Biomedicine 17 (4), 181–190. Xu, C., Brennick, M., Wootton, D., 2005. Image-based three-dimensional finite element modeling approach for upper airway mechanics. Conf Proc IEEE Eng Med Biol Soc 3, 2587–2590.