This study aims to quantify passive muscle stiffness of spastic wrist
flexors in stroke survivors using shear wave elastography (SWE) and to
correlate with neural and non-neural contributors estimated from a
biomechanical model to hyper-resistance measured during passive wrist
extension. Fifteen hemiplegic individuals after stroke with Modified
Ashworth Scale (MAS) score larger than one were recruited. SWE were used
to measure Young's modulus of flexor carpi radialis muscle with joint
from 0° (at rest) to 50° flexion (passive stretch condition), with 10°
interval. The neural (NC) and non-neural components i.e., elasticity
component (EC) and viscosity component (VC) of the wrist joint were
analyzed from a motorized mechanical device NeuroFlexor® (NF). Combining
with a validated biomechanical model, the neural reflex and muscle
stiffness contribution to the increased resistance can be estimated. MAS
and Fugl-Meyer upper limb score were also measured to evaluate the
spasticity and motor function of paretic upper limb. Young's modulus was
significantly higher in the paretic side of flexor carpi radialis than
that of the non-paretic side (
< 0.001). NC, EC, and VC on the paretic side were higher than the non-paretic side (
< 0.05). There was moderate significant positive correlation between the Young's Modulus and EC (
= 0.009) of the paretic forearm flexor muscle. Fugl-Meyer of the
paretic forearm flexor has a moderate significant negative correlation
with NC (
= 0.024). No significant
correlation between MAS and shear elastic modulus or NF components was
observed. This study demonstrated the feasibility of combining SWE and
NF as a non-invasive approach to assess spasticity of paretic muscle and
joint in stroke clinics. The neural and non-neural components analysis
as well as correlation findings of muscle stiffness of SWE might provide
understanding of mechanism behind the neuromuscular alterations in
stroke survivors and facilitate the design of suitable intervention for
them.
Introduction
Stroke is one of the major causes of long-term
disability worldwide and leads to motor and sensory impairments on upper
and lower extremity of survivors (
1,
2).
Spasticity, a neurological problem secondary to stroke, has a
significant effect on skeletal muscle and affects motor function and
quality of life (
3,
4).
Despite the high prevalence of spasticity in people with stroke, the
underlying mechanism remains poorly understood due to the confusion in
concept of spasticity. Alterations of spastic muscle might not simply
due to chronic stimulation or disuse, and literatures showed that
exaggerated reflexes and secondary changes in mechanical muscle
properties may have a major role in spastic movement disorder (
5).
Understanding the mechanical and neurophysiological characteristics of
spasticity may provide important clues to its intervention.
Previous studies involving a variety of methods studying
joint and tissue mechanics as well as muscle morphology offered clear
evidence to suggest that skeletal muscle tissue itself is altered under
spastic conditions (
6).
The mechanical property of spastic muscle could be assessed by imaging
technique such as ultrasound. Architectural parameters such as pennation
angles, fascicle lengths, and muscle thickness assessed by ultrasound
could quantitatively evaluate the morphological characteristics of the
muscle tendon complex (
7–
9). Shortened muscle fascicle length and smaller physiological cross section area were observed in patients with stroke (
7).
However, muscle function is not only related to its morphology
parameters but also the mechanical properties. Shear wave elastography
(SWE) tracks the propagation of the shear wave through the muscle tissue
using ultra-fast ultrasonic imaging and shear wave travels faster
through stiffer tissues (
10).
It is a new approach to provide a real-time quantitative metrics of
tissue material properties, including mechanical properties such as
stiffness of skeletal muscles in upper and lower limbs (
11–
15). Koo and coworkers found SWE was significantly correlated with passive muscle stiffness in both animal models (
16) and healthy subjects (
17).
Bouillard applied SWE and electromyography (EMG) to estimate individual
muscle force of healthy subjects and found a significant linear
correlation between shear elastic modulus and muscle force (
18).
These findings suggest that SWE may be a feasible way to quantify the
inherent strain-stress behavior of muscle after neuromuscular disease.
Published studies showed that shear wave speed and echo intensity were
higher in the relaxed paretic limb than in the relaxed non-paretic limb
of stroke survivors (
19–
21), multiple sclerosis (
22), cerebral palsy (
23–
25), and Duchanne muscular dystrophin (
26). However, previous studies of SWE on spastic muscle mainly focused on investigating relaxed muscles with a specific posture (
19–
21).
A systematic assessment of muscle stiffness in relation to joint angle
is needed to provide comprehensive information on muscle properties
alterations after pathology and how they relate to the ability to
conduct activities of daily living (ADL) such as feeding and dressing.
In clinics, physicians often examine the alterations in
muscle stiffness by passively stretching the spastic limbs and palpating
the spastic muscle. The outcomes of the assessments are subjected to
the experience of the clinicians. The sub-optimal evaluation method
adopted in clinics may due to the fact that quantitative assessment of
spastic muscle-tendon-joint is rare. Spasticity, defined as
hyper-resistance measured during passive rotation of a joint, is related
to neural and non-neural factors (
27–
29).
To separate and analyze individual components of this hyper-resistance
would lead to better understanding of spasticity and further assist the
selection of appropriate intervention (
28).
The NeuroFlexor (Aggro MedTech AB, Solna, Sweden) is a recently
developed motorized instrument which can passively extend the wrist
joint at controlled velocities (
30).
Based on the biomechanical modeling method, the recorded resistant
force can be separated into three contributors to the measured resistant
force: neural component (NC), elasticity component (EC), and viscosity
component (VC) (
31,
32).
It was found that the NC and EC increased in paretic side compared to
that from non-paretic side in stroke survivors (28.30). In addition, it
was reported that the NC decreased significantly after injections of
Botulinum toxin type A, but the passive components of EC and VC did not
change overtime (
31).
The NeuroFlexor (NF) has been also used to evaluate neural and
non-neural contributions to the wrist resistance in patients with
Parkinson's disease and cerebral palsy (
30–
33).
The validity of the NF-method has been demonstrated through the strong
correlation between the NC and the EMG activity of flexor carpi
radialis, both across all subjects and between-subject during the nerve
block test in the stroke survivors (
31,
34).
However, there is still controversial opinion on whether the neural
factor from central nervous system or the mechanical factor of
musculotendon system is a primary contributor to spasticity (
35).
In addition, there is limited information on whether the non-neural
components of resistance measured by the NF-method are clinically
relevant. For instance, it is unclear that whether non-neural components
are associated with intrinsic muscle mechanical properties.
The aim of the present study was to investigate the
feasibility and applications of combining SWE and NF to quantitatively
assess musculoskeletal properties alterations in spastic muscle joint
post-stroke. Understanding the delicate variations of muscles properties
during passive movement, and the neural and non-neural components that
contribute to joint resistance might facilitate the development of
targeted therapeutic regime to address spasticity for motor function
improvement.
Yan Leng1†, 
Zhu Wang
Ruihao Bian
Wai Leung Ambrose Lo
Ruoli Wang
Dongfeng Huang
Le Li