Limited fascicle shortening and fascicle rotation may be associated with impaired voluntary force-generating capacity in pennate muscles of chronic stroke survivors

BiLSTM-Based Joint Torque Prediction From Mechanomyogram During Isometric Contractions: A Proof of Concept Study

Quantifying muscle strength is an important measure in clinical settings; however, there is a lack of practical tools that can be deployed for routine assessment. The purpose of this study is to propose a deep learning model for ankle plantar flexion torque prediction from time-series mechanomyogram (MMG) signals recorded during isometric contractions (i.e., a similar form to manual muscle testing procedure in clinical practice) and to evaluate its performance. Four different deep learning models in terms of model architecture (based on a stacked bidirectional long short-term memory and dense layers) were designed with different combinations of the number of units (from 32 to 512) and dropout ratio (from 0.0 to 0.8), and then evaluated for prediction performance by conducting the leave-one-subject-out cross-validation method from the 10-subject dataset. As a result, the models explained more variance in the untrained test dataset as the error metrics (e.g., root-mean-square error) decreased and as the slope of the relationship between the measured and predicted joint torques became closer to 1.0. Although the slope estimates appear to be sensitive to an individual dataset, >70% of the variance in nine out of 10 datasets was explained by the optimal model. These results demonstrated the feasibility of the proposed model as a potential tool to quantify average joint torque during a sustained isometric contraction.

Effect of sit-to-stand-based training on muscle quality in sedentary adults: a randomized controlled trial

The aim of this study was to compare the effects of sit-to-stand (STS) training programs with 5 vs. 10 repetitions on muscle architecture and muscle function in sedentary adults. Sixty participants were randomly assigned into three groups: five-repetition STS (5STS), 10-repetition STS (10STS), or a control group (CG). Participants performed three sets of five or 10 repetitions of the STS exercise three times per week for 8 weeks. Before and after 8 weeks, all groups performed ultrasound measures to evaluate muscle thickness (MT), pennation angle (PA), and fascicle length (FL), and the five-repetition STS test to estimate the relative STS power and muscle quality index (MQI). After 8 weeks, both experimental groups improved MQI (40-45%), relative STS power (29-38%), and MT (8-9%) (all p < 0.001; no differences between the 5STS vs. 10STS groups). These improvements in both groups resulted in differences regarding the CG, which did not present any change. In addition, only the 5STS group improved PA (15%; p = 0.008) without differences to the 10STS and CG.This suggests that STS training is time-effective and low-cost for improving muscle function and generating adaptations in muscle architecture.

Characterization of the Structural and Mechanical Changes of the Biceps Brachii and Gastrocnemius Muscles in the Subacute and Chronic Stage after Stroke

The objective of this study was to characterize the changes of muscle tone, stiffness, and thickness of upper and lower limb muscles in stroke survivors. Forty patients with subacute or chronic stroke and 31 controls were included and measured using myotonometry (MyotonPRO), with multiple site assessments at muscle belly (MB) and musculotendinous (MT) locations of the biceps brachii and gastrocnemius muscles. Muscle thickness (ultrasonography) was obtained for each muscle. Upper and lower limb motor performance was evaluated with the Fugl−Meyer Assessment for Upper Extremity and the Functional Ambulance Category. Overall, muscle tone and stiffness were significantly higher at MT than at MB sites. Among stroke patients, differences between the paretic and nonparetic limb were found for the biceps brachii, with lower muscle tone, stiffness, and thickness of the paretic side (all, p < 0.05). There were weak to moderate correlations between mechanical (myotonometry) and structural (ultrasound) muscular changes, regardless of the post-stroke stage. This suggests that myotonometry and ultrasonography assess similar, although different, constructs and can be combined in the clinical setting. Their discriminative ability between the paretic and nonparetic sides and between participants with and without stroke differs depending on the muscle, the functional level, and the stroke stage.

Contribution of stroke-related changes in neuromuscular factors to gear ratio during isometric contraction of medial gastrocnemius muscle: A simulation study

BACKGROUND

It is not clear which neuromuscular factors are most closely associated with the loss of variable fascicle gearing after chronic stroke. The purpose of this simulation study is to determine the effects of stroke-related changes in key neuromuscular factors on the gear ratio.

METHODS

A modified Hill-type model of the medial gastrocnemius was developed to determine the gear ratio for a given muscle activation level and musculotendon length. Model parameters were then systematically adjusted to simulate known stroke-related changes in neuromuscular factors, and the gear ratio was computed for each change in the parameters. A Monte Carlo simulation was performed to understand which neuromuscular factors and fiber behavior-related parameters are most relevant to the loss of variable gearing. Dominance analyses were also conducted to quantify the relative importance of fiber behavior-related parameters on the gear ratio.

FINDINGS

The gear ratio decreases significantly with smaller pennation angle and with shorter optimal fiber length. In addition, muscle thickness and pennation angle at optimal fiber length appear to be the most important muscle architectural parameters. Dominance analyses further suggest that primary determinants of gear ratio include initial pennation angle, fiber rotation-shortening ratio, initial muscle thickness, and fiber rotation.

INTERPRETATION

Our findings provide insight that the pennation angle may play an important role for efficient muscular contraction, implying that maintaining muscle architecture and/or improving fiber/fascicle rotation could a key goal in rehabilitation interventions. Our findings will help us to better interpret altered gearing behavior in aging and pathological muscles.

Relative contribution of altered neuromuscular factors to muscle activation-force relationships following chronic stroke: A simulation study

The purpose of this study was to investigate the potential effects of key neuromuscular factors on muscle activation-force relationships, thereby helping us understand abnormal EMG-force relationships often reported in chronic stroke-impaired muscles. A modified Hill-type muscle model was developed to calculate muscle force production for a given muscle activation level and musculotendon length. Model parameters used to characterize musculotendon unit properties of medial gastrocnemius were adjusted to simulate known stroke-related changes in neuromuscular factors (e.g., voluntary activation and muscle mechanical properties). The muscle activation-force slope (i.e., muscle activation over force) was computed as a function of ankle joint angle. A Monte Carlo simulation approach was implemented to understand which neuromuscular factors are closely associated with the activation-force slope. Our simulations showed that a reduction in factors linked to voluntary activation capacity and to maximum force-generating capacity may be the primary contributors that increase the activation-force slope in dorsiflexed positions, and that a narrower active force-length curve appears to be the most significant factor that increases the slope in plantar flexed positions. In addition, our Monte Carlo simulation results demonstrated that an increase in the activation-force slope is strongly correlated with a reduction in voluntary activation capacity, in the maximum force-generating capacity, and in the active force-length curve width. These findings will help us to better interpret altered EMG-force relationships following chronic stroke.

Convergent Validity and Test-Retest Reliability of Multimodal Ultrasonography and Related Clinical Measures in People With Chronic Stroke

OBJECTIVE

To assess the test-retest reliability of diagnostic ultrasonography measurements of the bilateral biceps brachii (BB), brachial artery, medial gastrocnemius (MG), and popliteal artery in survivors of stroke and their convergent validity with related clinical comparators.

DESIGN

Cross-sectional study.

SETTING

All procedures were conducted in a university laboratory.

PARTICIPANTS

Sixty-five community dwelling adults (N=65; 26 women, 39 men) with an average age of 60.9±7.7 years and stroke duration of 5.7±3.9 years participated in this study.

INTERVENTIONS

Not applicable.

MAIN OUTCOME MEASURES

Measures of muscle structure (ie, thickness, cross-sectional area, fascicle length, pennation angle), stiffness, and intramuscular blood perfusion were conducted using B-mode, elastography and color flow Doppler ultrasonography modes, respectively. Convergent validity was assessed by examining correlations between ultrasonography measures and assessments of related constructs (ie, dynamic stiffness, isometric peak torque, spasticity, and systemic vascular function using myotonometry, dynamometry, the Composite Spasticity Scale, and the Ankle-Brachial Index, respectively). A 2-way random-effects intraclass correlation coefficient (ICC) model (ICC_2,3) was used to determine agreement between intersession measures among a smaller cohort of participants with stroke (n=20).

RESULTS

ICC estimates ranged from moderate to excellent for muscle stiffness (paretic: ICC=0.74-0.89; nonparetic: ICC=0.66-0.88), structure (paretic: ICC=0.87-0.99; nonparetic: ICC=0.81-0.98), and blood perfusion measures (paretic: ICC=0.74-0.84; nonparetic: ICC=0.73-0.88). Weak to moderate associations were found between myotonometry and elastography measures of the bilateral BB (r=0.29-0.52, P≤.05) and MG muscles (r=0.31-0.69, P≤.05). The correlations between elastography measures and spasticity scores for the paretic upper (r=0.35-0.63, P≤.05) and lower limbs (r=0.25-0.37, P≤.05) were also weak to moderate.

CONCLUSIONS

Elastography demonstrated mostly weak to moderate correlation with measures of stiffness using myotonometry as well as scores of paretic upper and lower limb spasticity. The results also indicate acceptable intersession reliability for muscle and vascular measures using several ultrasonography modalities among individuals with chronic stroke.

Changes in muscle-tendon unit length-force characteristics following experimentally induced photothrombotic stroke cannot be explained by changes in muscle belly structure

PURPOSE

The aim of this study was to assess the effects of experimentally induced photothrombotic stroke on structural and mechanical properties of rat m. flexor carpi ulnaris.

METHODS

Two groups of Young-adult male Sprague-Dawley rats were measured: stroke (n = 9) and control (n = 7). Photothrombotic stroke was induced in the forelimb region of the primary sensorimotor cortex. Four weeks later, muscle-tendon unit and muscle belly length-force characteristics of the m. flexor carpi ulnaris, mechanical interaction with the neighbouring m. palmaris longus, the number of sarcomeres in series within muscle fibres, and the physiological cross-sectional area were measured.

RESULTS

Stroke resulted in higher force and stiffness of the m. flexor carpi ulnaris at optimum muscle-tendon unit length, but only for the passive conditions. Stroke did not alter the length-force characteristics of m. flexor carpi ulnaris muscle belly, morphological characteristics, and the extent of mechanical interaction with m. palmaris longus muscle.

CONCLUSION

The higher passive force and passive stiffness at the muscle-tendon unit level in the absence of changes in structural and mechanical characteristics of the muscle belly indicates that the experimentally induced stroke resulted in an increased stiffness of the tendon.

Machine learning to extract muscle fascicle length changes from dynamic ultrasound images in real-time

BACKGROUND AND OBJECTIVE

Dynamic muscle fascicle length measurements through B-mode ultrasound have become popular for the non-invasive physiological insights they provide regarding musculoskeletal structure-function. However, current practices typically require time consuming post-processing to track muscle length changes from B-mode images. A real-time measurement tool would not only save processing time but would also help pave the way toward closed-loop applications based on feedback signals driven by in vivo muscle length change patterns. In this paper, we benchmark an approach that combines traditional machine learning (ML) models with B-mode ultrasound recordings to obtain muscle fascicle length changes in real-time. To gauge the utility of this framework for 'in-the-loop' applications, we evaluate accuracy of the extracted muscle length change signals against time-series' derived from a standard, post-hoc automated tracking algorithm.

METHODS

We collected B-mode ultrasound data from the soleus muscle of six participants performing five defined ankle motion tasks: (a) seated, constrained ankle plantarflexion, (b) seated, free ankle dorsi/plantarflexion, (c) weight-bearing, calf raises (d) walking, and then a (e) mix. We trained machine learning (ML) models by pairing muscle fascicle lengths obtained from standardized automated tracking software (UltraTrack) with the respective B-mode ultrasound image input to the tracker, frame-by-frame. Then we conducted hyperparameter optimizations for five different ML models using a grid search to find the best performing parameters for a combination of high correlation and low RMSE between ML and UltraTrack processed muscle fascicle length trajectories. Finally, using the global best model/hyperparameter settings, we comprehensively evaluated training-testing outcomes within subject (i.e., train and test on same subject), cross subject (i.e., train on one subject, test on another) and within/direct cross task (i.e., train and test on same subject, but different task).

RESULTS

Support vector machine (SVM) was the best performing model with an average r = 0.70 ±0.34 and average RMSE = 2.86 ±2.55 mm across all direct training conditions and average r = 0.65 ±0.35 and average RMSE = 3.28 ±2.64 mm when optimized for all cross-participant conditions. Comparisons between ML vs. UltraTrack (i.e., ground truth) tracked muscle fascicle length versus time data indicated that ML tracked images reliably capture the salient qualitative features in ground truth length change data, even when correlation values are on the lower end. Furthermore, in the direct training, calf raises condition, which is most comparable to previous studies validating automated tracking performance during isolated contractions on a dynamometer, our ML approach yielded 0.90 average correlation, in line with other accepted tracking methods in the field.

CONCLUSIONS

By combining B-mode ultrasound and classical ML models, we demonstrate it is possible to achieve real-time tracking of human soleus muscle fascicles across a number of functionally relevant contractile conditions. This novel sensing modality paves the way for muscle physiology in-the-loop applications that could be used to modify gait via biofeedback or unlock novel wearable device control techniques that could enable restored or augmented locomotion performance.

Quantitative Ultrasound to Assess Skeletal Muscles in Post Stroke Spasticity

Quantitative ultrasound (QUS) techniques such as pixel intensity, ultrasound strain, and shear wave elastography have made it possible to identify the echogenicity (brightness) and mechanical properties (stiffness) of normal and pathological tissues. These techniques can be utilized as an alternative diagnosis tool to assess post stroke spasticity. Current clinical assessment methods include the Modified Ashworth Scale (MAS) and the Modified Tardieu Scale (MTS), which can result in inconsistencies due to their subjective nature. QUS provides robust approaches to assessing muscle stiffness associated with post stroke spasticity. Computer-aided pixel count quantifies tissue echogenicity in grayscale image. A strain ratio in ultrasound strain imaging compares the stiffness and movement (lengthening or shortening) of a spastic muscle with nonspecific muscle. In addition, shear wave elastography provides the shear wave velocity of an affected muscle that directly associated with the muscle stiffness before and after treatment for spasticity. This article reviews the theory behind these aforementioned concepts and discuss the relations between QUS and skeletal muscles in post stroke spasticity.

Isolating the energetic and mechanical consequences of imposed reductions in ankle and knee flexion during gait

BACKGROUND

Weakness of ankle and knee musculature following injury or disorder results in reduced joint motion associated with metabolically expensive gait compensations to enable limb support and advancement. However, neuromechanical coupling between the ankle and knee make it difficult to discern independent roles of these restrictions in joint motion on compensatory mechanics and metabolic penalties.

METHODS

We sought to determine relative impacts of ankle and knee impairment on compensatory gait strategies and energetic outcomes using an unimpaired cohort (N = 15) with imposed unilateral joint range of motion restrictions as a surrogate for reduced motion resulting from gait pathology. Participants walked on a dual-belt instrumented treadmill at 0.8 m s^-1 using a 3D printed ankle stay and a knee brace to systematically limit ankle motion (restricted-ank), knee motion (restricted-knee), and ankle and knee motion (restricted-a + k) simultaneously. In addition, participants walked without any ankle or knee bracing (control) and with knee bracing worn but unrestricted (braced).

RESULTS

When ankle motion was restricted (restricted-ank, restricted-a + k) we observed decreased peak propulsion relative to the braced condition on the restricted limb. Reduced knee motion (restricted-knee, restricted-a + k) increased restricted limb circumduction relative to the restricted-ank condition through ipsilateral hip hiking. Interestingly, restricted limb average positive hip power increased in the restricted-ank condition but decreased in the restricted-a + k and restricted-knee conditions, suggesting that locking the knee impeded hip compensation. As expected, reduced ankle motion, either without (restricted-ank) or in addition to knee restriction (restricted-a + k) yielded significant increase in net metabolic rate when compared with the braced condition. Furthermore, the relative increase in metabolic cost was significantly larger with restricted-a + k when compared to restricted-knee condition.

CONCLUSIONS

Our methods allowed for the reproduction of asymmetric gait characteristics including reduced propulsive symmetry and increased circumduction. The metabolic consequences bolster the potential energetic benefit of targeting ankle function during rehabilitation.

TRIAL REGISTRATION

N/A.

Longer electromechanical delay in paretic triceps surae muscles during voluntary isometric plantarflexion torque generation in chronic hemispheric stroke survivors

Electromechanical delay (EMD) is the time delay between the onset of muscle activity and the onset of force/joint torque. This delay appears to be linked to muscular contraction efficiency. However, to our knowledge, limited evidence is available regarding the magnitude of the EMD in stroke-impaired muscles. Accordingly, this study aims to quantify the EMD in both paretic and non-paretic triceps surae muscles of chronic hemispheric stroke survivors, and to investigate whether the EMD is related to voluntary force-generating capacity in this muscle group. Nine male chronic stroke survivors were asked to perform isometric plantarflexion contractions at different force levels and at different ankle joint angles ranging from maximum plantarflexion to maximum dorsiflexion. The surface electromyograms were recorded from triceps surae muscles. The longest EMD among triceps surae muscles was chosen as the EMD for each side. Our results revealed that the EMD in paretic muscles was significantly longer than in non-paretic muscles. Moreover, both paretic and non-paretic muscles showed a negative correlation between the EMD and maximum torque-generating capacity. In addition, there was a strong positive relationship between the EMD and shear wave speed in paretic muscles as well as a negative relationship between the EMD and passive ankle joint range of motion. These findings imply that the EMD may be a useful biomarker, in part, associated with contractile and material properties in stroke-impaired muscles.

Loss of variable fascicle gearing during voluntary isometric contractions of paretic medial gastrocnemius muscles in male chronic stroke survivors

KEY POINTS

Maximum fascicle shortening/rotation was significantly decreased in paretic medial gastrocnemius (MG) muscles compared to non-paretic MG muscles. The fascicle gear ratio on both sides decreased as the ankle became dorsiflexed, but the slope of the fascicle gear ratio over ankle joint angle was significantly lower on the paretic side. The side-to-side slope difference was strongly correlated with the relative maximum joint torque and with the relative shear wave speed, suggesting that variable gearing may explain muscle weakness after stroke.

ABSTRACT

The present study aimed to understand variable fascicle gearing during voluntary isometric contractions of the medial gastrocnemius (MG) muscle in chronic stroke survivors. Using ultrasonography, we characterized fascicle behaviour on both paretic and non-paretic sides during plantarflexion contractions at different intensities and at different ankle joint angles. Shear wave speed was also recorded from the MG muscle belly under passive conditions. Fascicle gear ratios were then calculated as the ratio of muscle belly shortening velocity to fascicle shortening velocity, and variable fascicle gearing was quantified from the slope of gear ratio vs. joint angle relations. This slope was used to establish associations with maximum joint torques and with shear wave speeds. At all measured angles, we found a significant reduction in both maximum fascicle shortening and maximum fascicle rotation on the paretic side compared to the non-paretic side on our stroke survivor cohort. The fascicle rotation per fascicle shortening on the paretic side was also significantly smaller than on the non-paretic side, especially at plantarflexed positions. Furthermore, the fascicle gear ratio on both sides decreased as the ankle became dorsiflexed, but the change in the fascicle gear ratio was significantly lower on the paretic side. The side-to-side difference in the gear ratio slope was also strongly correlated with the relative maximum joint torque and with the relative shear wave speed, suggesting that variable gearing may explain muscle weakness after stroke. Further studies are needed to investigate how muscular changes after stroke may impede variable gearing and adversely impact muscle performance.