Institute of Biomedical Engineering. National Yang-Ming University.
No.155, Sec. 2, Linong St., Beitou District, Taipei City 112, Taiwan (R.O.C.).
Background: Muscle properties changes early after stroke. Accurately characterizing and quantifying the stiff muscle is clinically important to better understand the altered muscle function and movement control.
Objective: The aim of this study was to investigate the feasibility of sonoelastography to determine the muscle stiffness and the echotexture in poststroke. We also tested the relationship among sonoelastography findings, muscle echotexture features and functional performance in the spastic muscle.
Methods: A total of 21 men with subacute stroke were studied. The intrinsic stiffness of biceps brachii muscles (BBM) on both arms were assessed at rest by shear wave velocity and echotexture features (entropy & energy) were extracted by Local Binary Pattern (LBP) of ultrasound imaging. The scanning images of BBM were acquired in both the transverse and the longitudinal planes. The Fugl-Meyer Assessment (FMA) and Functional Independence Measure (FIM) were used to assess the functional performance of upper arm.
Results: The shear wave velocity was significantly faster in paretic BBM, compared to non-paretic BBM in the transverse and the longitudinal planes. The echotexture (entropy & energy) was more in-homogeneous in the paretic BBM than in the non-paretic side on both scanning planes. The shear wave velocity was negatively correlated to entropy (r = – 0.44, P = 0.04) and energy (r = 0.49, P = 0.02) in the longitudinal plane. The energy was correlated to FMA (r = – 0.46, P = 0.03), FIM (r = – 0.55, P < 0.01) in the longitudinal plane, and duration from stroke onset (r = 0.49, P = 0.03), age (r = – 0.53, P = 0.02) in the transverse plane.
Conclusion: The echotexture of LBP is capable to be a useful tool for quantitative assessment of the spastic BBM in patients with early stroke.
shortcut of Ginko：cmd + M
= / -:Focus mode
full screen: F
#: header (#, ##, ### 中間空一格)
tag: #tag (#連在一起)
斜體： ltalic text . cmd+I
列點： item. cmd+I then space
圖片：cmd + shift + I
新增卡片：cmd + arrow key
確認：cmd + enter
刪除：cmd + delete
Background: The aim of this study was to identify the echotexture of paretic muscle in stroke patients.
One way to describe the muscle properties is “stiffness”
Methods to assess stiff, xxx
Shear wave velocity was used to measure the stiffness of biceps brachi muscle and echotexture was extracted by local binary pattern. The differences between paretic and non-paretic BBM were evaluated. The correlations among the stiffness, echotexture and functional performance were also tested.
1. The echotexture (entropy & energy) is different between paretic and non-paretic BBM in both plane
2. The shear wave velocity is faster in paretic BBM. in transverse and longitudinal planes for stroke patients
3. The energy/entropy is correlated to functions (FIM/fugl meyer) and onset duration, age. But
none of echotexture is correlated to swv
4. The shear wave velocity is negatively correlated to ENT/ENG in longitudinal plane -> concentration theory. WHY transverse no correlated?
5. THe R of LBP is different in both plane in both plane? The optimal radius of R is 3
*The echotexture extracted by local binary pattern may be a useful diagnositic tool for assess and follow-up the muscle property changes after stroke.
Spasticity (hyperactiviity), muscle weakness (hypoactivity) and joint contracture (immobilization) are often seen consequences poststroke and cause considerable long-term motor impairments to stroke survivors.
Over time, material properties of muscles in the impaired limbs can also change gradually, further disrupting motor function, and adversely impacting the stroke survivor’s quality of life.
Subsequent to these impairments, morphologic changes in the architecture of the paretic muscles often occur, which affect the muscles’ functions. Previous studies have revealed a reduction in muscle volume, a shortening of muscle fiber, and a reduction in the number of motor units in the paretic muscles in people after stroke. `These muscle deformations were highly related to syndromes of paretic muscles, such as muscle weakness, spasticity, and contracture.
Evaluating muscle property variations in people after stroke is clinically important for both diagnosis and rehabilitation treatment.
Muscle property changes includes IMF, EMC, connective tissue accumation. Recently, sonography is using b mode and EI to assess these changes. But EI can not be compared. More recent advanced imaging, SW is an alternative tool to measure the muscle stiffness.
Spasticity lead to chagnes in muscle including muscle stiffness and muscle property change.
Weakness lead to loss fiber elasticiity.
Joint contracture leads to shorten length of muscle fiber and joint stiffness.
“The in vivo assessment of the biomechanical properties of the skeletal muscle is a complex issue because the muscle is an anisotropic, viscoelastic and dynamic medium” (Gennisson et al 2010:789)
Examining the mechanical properties of muscle is important in monitoring the stage of the pathologic processes of muscles and for assessing efficacy of therapeutic interventions
One of the way to descript the material property is stiffness.
Past studies try to use MAS/myotonometr etc., to measure the muscle stiffness
the Ashworth scale and modified Ashworth scale remain low-sensitivity instruments for distinguishing between soft tissue and neural contributions to hypertonia
-> however it is not reliable / not handy. -> lack of diagnostic biomarkers
Paet studies use US to assess the muscle quality, but operator dependant. [xxx]
More recently, Echotexture features was used to test the muscle quality to minimize the operator factors. Interpreted as the uniformity of the muscle pattern, provide further information about muscle property, beside the echo intensity which is highly dependent on the ultrasound scanner setting (plane, anisotropy) and wave energy scattering i.e. the distribution of the pixel intensity.
Recently SWE’s speed is related to stiffness (young’s modulus) -> high SWV is high stiff, SWE can assess muscel mechanincal properties (muscle stiffness) [xxx]
B mode can not provide muscle stiffness information which might be related to spasticity
Many studies showed that stiffness increased in spastic muscles due to changes in fiber type transformation. Morphometric and
histochemical investigations show changes in mechanical muscle-fiber properties that might contribute to spastic muscle tone [7-12]. Changes in biomechanical conditions of a muscle might also have an important effect on the spasticity in people with stroke.
Shearwave (SW) ultrasound elastography,which builds upon traditional elastography, allows quantitative in vivo measurement of tissue material properties (Bercoff et al., 2004). Using the same acoustic radiation forces as B-mode ultrasound, we use the SuperSonic Shear Imagine (Bercoff et al., 2004), a method that uses multiple ultrasound push
beams to induce the SWs and subsequently, to measure the SW speed in muscle.
The speed of these SWs is related to material properties, such that SWs travel faster through stiffer tissues.
Sonoelastography is based on the calculation of Young’s elastic modulus, a physical quantity measuring stiffness
This technique has emerged as a reliable method to estimate material properties in a variety
of tissues (Bouillard et al., 2012), including muscle (Bouillard et al.)
2012; Eby et al., 2013; Lacourpaille et al., 2012).
Recently, SW speed in the biceps brachii muscle of the paretic side was found to be
on average 69.5% greater than the non-paretic side in stroke survivors
(Lee et al., 2015)
Shear waves are tracked by pulse-echo ultrasound and can be used quantitatively to calculate the tissue modulus (ie, stiffness); as the stiffness of underlying tissue increases, shear-wave speed increases.
The stiffness is correlated to muscle tone and muscle property changes results in muscle architexture change after a stroke. Although SWV measure muscle in a in-directly way.
Thus, considering physical activity leading to angiogenesis and muscular fiber changes, and inactivity leading to sarcopenia and fat infiltration, we could expect that muscle plasticity would induce stiffness changes. However,
SWE did not reveal the quantitative stiffness changes expected in relation to the specific muscle histology of samples extracted from females or males, athletes, juniors or seniors.
Little known of the echotexture of stiff muscle. and echotexture related to arm functions.
Early Intervention reference, treatments follow up
Under the assumption that pathological muscle tissue appears to be diffuse and unstructured, a texture-based analysis seems adequate for detecting XXX in ultrasound images.
1) in the past study, us was used to test the muscle quality changes, (who check the stiffness by traditional US?) what do they find? I persume that muscle get stiffer, the velocity get higher, and the echotexture changes as well.
2) the resting muscle tone (combine neurological plasticity, and properties changned). So echotexture might help to identify the muscle properties part rather than the neurological part. So, Stiff muscle = mas + muscle property.
(muscle stiffness = tone + biomechanical factor)
隨著影像技術的進度，SWV 也是評估肌肉痙攣與硬度的一種方式，WU&xxx嘗試(check my summary notes)，雖然SWV可以提供波速純量，但是SWV較受人為，壓力影響還有anisotropy的因素，在者波素與紋理特徵也沒有清楚了解，因此，本研究的目的即是探討痙攣肌肉紋理特徵與硬度的關係，並分析紋理/硬度與功能的關聯性。
A total of Twenty-two stroke patients finished the sonoelastograpic measurements of biceps brachii muscle and clinical evaluations. The demographic data and functional assessments are summarized in Table 1.
Sabrina S.M. Lee, 2015 : difference of swv correlated to difference of echo intensity at biceps (check the Fig 5) and the explainations in content.
ENT/ENG可以區辨痙攣BBM)The echotexture of spastic muscle is defined by high order statistical textual analysis with local binary patterns in energy and entropy.
Muscle fiber in shirt position, sacromere reduce, architecture change makes fiber shorten. The sound wave reflect environment changed, so the echo texture change(what changes?)
Normal muscle scan in longitudinal plane, the perusing is clear to seen, but in myo or neuropathy become blue it is related to anisotropy
Compared to Echo Intensity
BB muscle for the transverse and longitudinal scans (Table 1)
had similar mean values, standard deviation and CVs, but different
ICCs between probes orientations, being the transverse
one more reliable (r = 0!84) than the other (r = 0!39). (da Silva, 2017)
Unfortunately, these reviews target health care providers
with a strong background in ultrasound physics and provide limited discussion of the clinical application and significance of ultrasound elastography with respect to muscle. Thus, they are of little assistance to the typical physical medicine and rehabilitation physician seeking to improve clinical practice by adding ultrasound elastography. Many rehabilitation strategies are aimed at changing the mechanical properties of muscle.
Thus, considering physical activity leading to angiogenesis
and muscular fiber changes, and inactivity leading to
sarcopenia and fat infiltration, we could expect that muscle
plasticity would induce stiffness changes. However, SWE
did not reveal the quantitative stiffness changes expected in relation to the specific muscle histology of samples extracted from females or males, athletes, juniors or seniors.
Muscle stiffness and muscle property, as indicated by SW speed and echotexture, were changed in stroke-impaired muscle at rest. These findings highlight the potential for echotexture features extracted by LBP as a tool for both investigating the fundamental changes in stroke-impaired muscle, and for evaluation of muscle mechanical properties as part of clinical examination.
“Proper evaluation of muscle properties is important in monitoring the progression of individuals during rehabilitation therapy, in making appropriate clinical decisions, for planning optimal treatment, and for assessing the efficacy of therapeutic interventions.” (Chuang et al 2012:533)
stroke is most common NMD -> impairement = neural (primary) + muscle property (secondary) = Quantify muscle property is challenge.
Muscle properties changes are xxx -> impact movement.
Measuremen are XXX -> however these prior methods have multiple limitations including
Invaluable information of stiffness mechanism at cellular and fiber livel
My results showed an odd pattern. the swv is negatively correlated to entropy. But in my findings, the paretic arm of entropy is higher than non-paretic arm. It is totally conflicts.
the more harder (SWV) the muscle, the higher entropy has. It should be positive correlation between swv and entropy. But my result did not meet this finding when perform the spearaman correlation test. It has strong negative correlation between them.
Stroke is long term disability. Impairments include XXX. These impairements lead to muscle properites change over time -> impact QoL.
Muscle stiffness, past studies results -> However, methods were made indirectly.
Quickly quantify changes in muscle stiffness of specific muscles in a clinical setting remains a challenge.
Recently, SWV have investigated the stiffness on XXX. High speed is in spastic muscle. -> indirectly estimate stiffness.
Accordingly, we sought to determine XXX. We also assessed the echotexture of BBM and correlated our estimates of material properties with major clinical assessments in cluding, FIM FMA.
IMF ECM connetive tissue acculmation lead to further fiber shorten, visco-elasticity thxithplaticity, joint contrature. Therefore, the architecture of muscle changes and more straight the echotexture is changed,
Histolopathological studies have demonstrated a generalized increase in extracellular connective tissue in spastic muscles. It is known that increased connective tissue in an immobilized and contracted muscle reduces its compliance (Stecco et al 2014:121)
muscle mechanical properties could provide clinically viable information to follow the effects of neuromuscular disorders or potential improvements due to various treatments.
The disruption of normal muscle architecture caused by the infiltration of fat and the development of fibrosis has been reported to increase reflections of the ultrasound beam, resulting in an increased echo intensity of spastic muscles. Based on this, the development of changes in the muscle structure has been reported to influence the echotexture.
MAS might not be as sensitive in detecting muscle stiffness.
Enhanced echo intensity (EI) on ultrasonography images of skeletal muscle is believed to reflect muscle quality.
EI is based on the histogram of the greyscale pixels of the ROI. The histogram provides a unidimensional analysis and classifies the tissue as hyperechoic or hypoechoic.
To overcome XXX limitations, image texture anslsysis can be used as it is a mathematical proceduure that takes into accout the two-dimensional distribution of pixels in the ROI and can be obtained by several statistical techniques XXXX for US image texture characterization.
However, B-mode ultrasonography does not provide information on muscle stiffness, which may be associated with spasticity. Use of ultrasound elastography is a recently developed, ultrasound-based method to evaluate tissue elasticity.
Because the texture of a tissue image reflects its scatterers’ properties, the entropy can be used to characterize the texture of an input image.
(Turo, 2013) use ENT a measure of heterogeneity of the tissue for differentiating tissue problem such as…
SWE has been reported to be a reliable tool for detecting muscle stiffness. For example, XXXX. However, their role as diagnostic or progression biomarkers is unknown, because of the high inter-individual variability in structural parameters. Therefore, it would be XXX to find a MUS biomarker not influenced by factors other than disease factors. Echotexture has been previously reported to characterize XXX.
“Quantitative elastography might help by providing a parameter that could influence the diagnosis but it could also provide a way to monitor the effectiveness of a treatment by quantifying the changes of the muscle mechanical state.” (Gennisson et al 2010:790)
Hypothesis, right and left are different, transverse and the longitudinal plane but is different
The aim of the current study is to quantify differences in shear wave velcoity and echotexture between subjects with stroke from ultrasound using Local binary pattern image analysis methods.
We hypothesize that simple grayscale parameters estimating quantitative muscle
echotexture parameters derived from LBP-based images would be different between paretic and non-paretic BBM.
The study protocol was approved by the Research Ethics Committee of a medical center hospital and written informed consent was obtained from all subjects.
All subjects were recruited from a rehabilitation ward in a medical center. The Inclusion criteria were (i) age >30 years, (ii) absence of cervical radiculopathy symptoms or other central or peripheral neurologic deficits in the upper limbs, (iii) bilateral elbow range of motion flexion within 0–145 degrees, and (iV) onset duration less than 3 months. Exclusion criteria were (i) history of botulinum toxin injection; and (ii) history of upper limb surgery.
All subjects were placed supine and instructed to relax his examed elbow during ultrasound imaging capture. A low-temperature premade arm splint was used to standardize the arm position and to minimize the activity of recorded muscles. The shoulder was placed so that the humerus was abducted 45°, and the elbow positioned at 90°.
Give a photo for demostration
B-mode ultrasound scanning images and SWV values were captured from both elbows using an Acuson S2000 ARFI Ultrasonography System (Siemens, Munich, Germany) equipped with a 7~9 MHz linear transducer (9 L4, Siemens).
To ensure the applied proper pressure on the transducer during SWV acquisition, the “Quantify index” above 80 is needed for high reliable test value.
All scans of biceps brachii muscle were acquired by placing the transducer at
two thirds of the distance from the acromion to the antecubital crease (Molinari et al 2015:2521)
“The probe was oriented parallel to the muscle fibers to obtain SWV measurements in the longitudinal axis and rotated 90 for SWV measurements in the transverse axis.” (Wu et al 2017:1106)
“BB US images were acquired at 60% of the distance between the posterior ridge of the acromion and the olecranon of the right arm, while the subject was seated with his arms relaxed on their respective sides (Matta et al., 2011).” (da Silva Pereira Júnior et al 2017:85)
The region of interest (ROI) was set to the mid portion of the BBM thickness for measuring the BBM stiffness. The scanned images were then transferred to a computer in the Digital Imaging and Communications in Medicine (DICOM) format for further extracting the echotextural features.
ROI size of 142 x 142 pixels for BBM on an 8-bit gray-scale was selected at which muscle SWV was measured.
Sonoelastography was performed in both transverse
and longitudinal planes. In each plane, three measurements of SWV and ROI were acquired from bilateral biceps brachii and averaged for further statistical analysis.
“Five measurements of SWV were acquired from bilateral biceps brachii in each elbow posture and averaged for further statistical analysis” (Wu et al 2017:1107)
The unaffected contralateral side was examined at the same time under identical conditions and served as normal image control
please refer to Hirvasniemi, 2017, bone texture
“A Local Binary Pattern (LBP) (Ojala et al., 2002) is a local texture operator with powerful discrimination, low computational complexity, and less sensitivity to changes in illumination. The LBP operator is obtained by evaluating the binary difference between the gray value of a pixel x from the gray values of its P neighborhood placed on a circle radius of R” (Nanni et al 2010:7891)
“To calculate the LBP operator the local binary difference between the gray value of a pixel x and the gray values of P pixels in a local neighborhood of x placed on a circle of radius R is evaluated. To obtain a rotation invariant descriptor , P 1 bitwise shift operations on the binary pattern are performed, and the smallest value is selected. A pattern is defined “uniform” if the number of transactions between “0” and “1” of the sequence is less or equal to two, with the number of different types of uniform patterns that can occur being P + 1. To describe a given image, the histogram of dimension P + 2 is extracted. It contains the occurrence of the P + 1 types of uniform patterns, and the number of non-uniform patterns.” (Nanni et al 2010:118)
The US image with irregular informative texture. This can be quantitfied by texture features like local binary pattern.
Two features were extracted from each ROI.
These features are based on statistical and transformation on the ROI.
The textural descriptor is LBP. and the LBP features are ENT / ENG.
“Entropy: The entropy is a measure of average information in a block.” (Brehar and Nedevschi 2013:119)
“The entropy is high when the local binary patterns or the gradient magnitudes are distributed among many values.” (Brehar and Nedevschi 2013:119)
“Energy: In our work we define energy as a measure that shows how the local binary patterns or the gradient magnitudes are distributed on different bins.” (Brehar and Nedevschi 2013:119)
“The energy is low when the number of uniform patterns or the number of gray levels of the gradient magnitude are high. The energy is maximum when we have only one LBP pattern or only one gradient magnitude value in the block.” (Brehar and Nedevschi 2013:119)
The local binary pattern image was then calculate the entropy and energy. The rodness of image is high for entroy.
Entropy: The entropy is a measure of average information in a block. We use the following formula for computing the entropy of a block:
H = - p(b)*log2(p(b))
The entropy is high when the local binary patterns or the gradient magnitudes are distributed among many values.
Energy: In our work we define energy as a measure that shows how the local binary patterns or the gradient magnitudes
are distributed on different bins. We define the energy of a block as:
E = (p(b))2
The energy is low when the number of uniform patterns or the number of gray levels of the gradient magnitude are high.
The energy is maximum when we have only one LBP pattern or only one gradient magnitude value in the block.
“We considered P equal to 24 pixels and R equal to 3 pixels, in order to consider a relatively large neighborhood.” (Molinari et al 2015:2532)
BBM spasticity was evaluated at the paretic arm by the Modified Ashworth Scale (MAS) and the Tardieu Scale (TS).
The MAS is a simple evaluative method to assess the spasticity with the six levels scoring. The 0 = no increase in muscle tone; 4 = affected part(s) rigid in flexion or extension.(Pandyan et al. 2003)
measures resistance during passive soft-tissue stretching and is used as a simple measure of spasticity. Scoring (taken from Bohannon and Smith, 1987):
TS evaluates the spasticity angle (the angle at which muscle reaction occurs at 3 velocities: as slow as possible (V1), falling under gravity (V2), and as fast as possible (V3).
Fugl Meyer Assessment
Functional Independent Measure (FIM)
If the p value is greater than alpha value:0.05, the null hypothesis is accepted
So the tested data is normality.
Wilcoxon signed rank paired t-test were uesd to test the difference of variables bewteen paretic and non-paretic BBM.
Wilcoxon signed-rank test was used for comparison between the hardness with the use of the two types of
As the Shapiro – Wilk test for normal distribution of the data
failed, non-parametric tests were used. The Wilcoxon signed
rank sum test was used to compare muscle echo intensity and
muscle and subcutaneous layer thickness between the two sides
(dominant versus non-dominant).
The relationships between the actual and
measured hardness were examined by the least square method.
SPSS® statistics v. 22 (IBM Corp., Tokyo, Japan; formerly SPSS Inc., Chicago, IL) was applied for statistical analysis. Values of p < 0.05 were set to indicate significance.
The Spearman rank correlation analysis was used to test for linear correlations between echo intensity and thickness measurements
To test the differences among R size using “Fridaman test”
=> nonparametic k-related t test
Figure 2 showed that the spastic BBM had signicantly higher shear wave speed (P < 0.001) in both the transverse and the longitudinal planes.
The ENT_r2 was higher for paretic BBM in the Longitudinal plane (p = 0.05). The ENG was non-significant difference between paretic and non-paretic BBM in the lingitudinal plane.
In the transverse plane, The echotexture of paretic BBM was significantly in-homogeneous than non-paretic BBM in ENT_r1 and ENG_r1.
The size of radius, less than 3 is enough to differintiate the echotexture for paretic and non-paretic BBM.
To our’s knowledge, no prior work has been reported on the image texture feature based quality assessment of paretic muscle ultrasound images using local binary patterns.
In this paper, the information of image texture features are analyzed using LBP with entropy and energy, and also correlated with muscle stiffness and functional performance.
Recently, SW speed in the biceps brachii muscle of the paretic side was found to be on average 69.5% greater than the non-paretic side in stroke survivors (Lee et al., 2015)
Theoretically, increased resistance (
Stiffness) to passive movement can result from neural factors such as spasticity, or peripheral factors such as changes in the mechanical properties of the muscles (for example, as a result of muscle contracture)
Entropy is a term derived from thermodynamics and refers to the quantity of energy that is permanently lost to heat (“chaos”) every time a reaction or a physical transformation occurs. The term is used to non-technical speech to mean “irremediable chaos or disorder”.
Energy (ENG) is the opposite of entropy and represents orderlines in the image.
-> In the past study, swv is positively correlated to echo intensity. It represented that the more fat infiltrated into muscle, the elasticity of muscle is lower. In this study, we choose echotexture to test the muscle properties in early stage after stroke. The onset duration get longer, it is reasonable supposed that more fat will infiltrate to muscle. According to concentration theory, the more fat in muscle, the texture of muscle get homogeneous (cite??) therefore the ENT become smaller.
“A combination of denervation, disuse, inflammation and remodeling contributes a complex pattern of muscle tissue phenotype change and atrophy that may factor into a decrease in muscle mass and tone” (Gao et al 2018:8)
“Acharya et al. reported that two powerful descriptors of the LBP image are the energy and entropy of the LBP distribution” (Molinari et al 2015:2524)
In our results, the paretic can prove it, the echotexture is more in-homo than health side.
when the muscle get stiffer, the density of muscle property become high. As the fiber become more, the echo internsity is whiter. then the ENT gets smaller (homogeneous).
To miminize the operator factor and age, gender, this (ENT/ENG) could be a follow-up biomarker.
Additionally, most SWE techniques
assume that the underlying tissue is isotropic, elastic, and locally homogenous—such as that of breast, liver, or thyroid. Muscle, however, is anisotropic: the mechanical properties along muscle fibers differ from those across muscle fibers. This anisotropy requires orientation of the transducer for all SWE techniques to be longitudinal to muscle fibers in order to achieve accurate and reliable measurements. Despite the anisotropy, the shear modulus (a stiffness measure that assumes isotropy) measured from shear-wave speed displays good agreement with the Young modulus (a stiffness measure that assumes isotropy and incompressibility) throughout the range of normal physiologic tension of skeletal muscle. In the medical literature, the shear modulus (or shear elastic modulus) and the Young modulus have both been used in reporting outcomes.
muscle stiffness only when the muscle is not stretched. All significant differences and correlations were detected only when the SWV of the biceps brachii muscle was evaluated in the longitudinal axis, not in the transverse axis. The mechanical properties along fibers differ from those across fibers (Brandenburg et al. 2014) (i.e., muscle is anisotropic), so the ultrasound transducer should be oriented longitudinally to the muscle fibers to obtain SWV values with functional relevance (Gennisson et al. 2010). Moreover, Dorado Cortez et al. (2016) reported better reproducibility of muscle SWV measurements in the longitudinal plane than the transverse plane. 2.98 6 0.11 m/s, p 5 0.529). Paretic-side SWV in the longitudinal axis was positively correlated with stroke duration at 90 elbow flexion, but not at 0 . Paretic-side SWV in the longitudinal axis correlated positively with MAS and MTS at both 90 and 0 , but the correlation coefficients were higher at 90 . There was a negative correlation between pareticside SWV in the longitudinal axis and STREAM score (Table 3). In contrast, there were no correlations between paretic-side SWV in the transverse axis and post-stroke duration, MAS, MTS and STREAM score (all p . 0.2). The ICC of inter-rater reliability for SWV measurements was 0.768 (0.373–0.927, ‘‘excellent’’) in the longitudinal axis and 0.552 (0.002–0.846, ‘‘good’’) in the transverse axis.
The in vivo assessment of the biomechanical properties of the skeletal muscle is a complex issue because
the muscle is an anisotropic, viscoelastic and dynamic medium.
“The resulting pattern is captured as an 8-bit binary number representing one of 256 distinct known patterns. Then, the histogram is computed for the transformed image and considered as a texture-descriptor. The LBP may fail in many cases for anisotropic phenomena since it is more complex than natural textures. Anisotropy is characterized by the global and privileged directions of the structure. The 2D local patterns are less sensitive to such characteristics, because they encode only the frequency of local structures regardless of their global orientations.” (Houam et al 2014:185)
“Depending on the orientation analyzed, results are different, proving that non-uniform changes due to osteoporosis induce variations in the degree of anisotropy.” (Houam et al 2014:192)
“All commercially available SWE systems are based on the prerequisite that soft tissues are purely elastic, incompressible and isotropic. First, the major technical parameter that influences stiffness measurement is the anisotropic physical properties of the skeletal muscle. The tissular organization of skeletal muscle, which comprises a parallel arrangement of myofibrils, muscular fibers, collagen and elastic fibers, and fascicles, confers anisotropic, in particular orthotropic properties (which are a subset of anisotropic properties that differ along the three orthogonal axes) to the skeletal muscle. These orthotropic physical properties are responsible for the fact that shear waves travel faster along the direction of the fibers than they do when perpendicular to them [19, 21] (Fig. 1). This has a number of consequences. First, stiffness measurements are sensitive to the angle between the probe axis and the orientation of the muscular fibers. Shear modulus measurements using SWE are correlated with Young’s modulus only if the probe is oriented parallel to the muscle fibers. Another consequence is the difficulty assessing meaningful results in muscles with complex anatomy. Multipennate, conic, triangular or fusiform anatomy, which yieldsBmultiorientation^ fibers, introduces a technical difficulty in visualizing the orientation of fibers.” (Creze et al 2017:2)
主要影響的技術問題就是肌肉是”非同質性物質”，造成方向依賴，無法使用於複雜的組織。 (note on p.2)
The LBP is good to distingish the muscle type by echotexture analysis. therefore, may used in sarcopenia, the structure changes in the extracellular matrix (ECM), such as an increase in the collagen concentration, may be associated with the EI of skeletal muscle. It is also possible that the structural and biochemical changes in skeletal muscle ECM contribute to aging-related loss of muscle function (eg, impaired in force generation and increased stiffness).” (Watanabe et al 2013:996)
Ultrasound (US) is an important tool for diagnosing of many musculoskeletal tissue conditions. ‘Image texture analysis can be used to characterize the types of tissue. The morphological and textural characteristics of US images are commonly used for qualitative diagnosing (Rahbar et al., 1999; Huber et al., 2000). Quantitatively, the texture image analysis is based on the spatial variation of the pixel intensity and several parameters can be statistically defined by the probability of the pixels’ greyscale values. One example is the Entropy / Energy, which is the ruddiness and homogeneity of an image.
Shear wave velocity
“US elastography is commercially available and can be employed in clinical settings. The elastic properties of muscular tissue are one way of diagnosing and evaluating degenerative myopathies, to determine the best rehabilitation program for stroke patients, and diabetes (13). It i” (Niitsu et al 2011:104)
When muscle architecture and muscle hardness are
measured concurrently, the B-mode ultrasound image and the ultrasound elastogram are acquired at the same time and in the same imaging plane. In addition, the time required to obtain a B-mode ultrasound image concurrently with an ultrasound elastogram is shorter than that required to obtain them separately. Thus, this combined technique offers an efficient way to objectively assess muscle architecture and muscle hardness, which are both indicators of muscle condition.
In ultrasound shear-wave elastography, shear modulus is calculated from the propagation velocity of shear wave inside tissues, which can be spatially variable when tissues are made of different elasticities.
In its current application to muscles, the spatial average of shear modulus distribution across a region of interest (ROI) of a muscle has been determined using a commercially available ultrasound device with included software. The problem is that a simple spatial average of shear modulus distribution across an ROI in a muscle is not always appropriate for assessing contraction level because it can include not only shear modulus of contractile tissues (i.e. muscle fibers) that produce active force with contraction, but also non-contractile tissues (NCT, i.e., skin, adipose tissue, and connective tissue) that can be distributed within the contractile tissue region. Still, the impact of NCT inclusion in the image analysis on the assessment of muscle contraction intensity is unknown. (NCT: non-contractile tissue)
Shear modulus is the biomechanical parameter used to characterize stiffness” (Creze et al 2017:2)
It is based on Hooke’s law, a relationship among strain, stress and elasticity: E = d, where the s elastic or Young’s modulus, is measured in kPa; s is the stress d or external force; d is the strain or deformation” (Creze et al 2017:2)
Because elasticity is a critical determinant of
muscle performance and force; hence, its assessment in vivo can help to improve the understanding of muscle functions” (Creze et al 2017:2)
The elastic properties of muscular tissue are one way of diagnosing and evaluating
degenerative myopathies, to determine the best rehabilitation program for stroke patients, diabetes (Niitsu et al 2011:104) , pain, such as tension-type headache, muscle belly swelled up and macro-architecture of the muscle changed. (Niitsu et al 2011:104)
US elastography demonstrates color map of muscle hardness and can be applied to all striated muscles.