Relationship between lower limb muscle coordination and knee flexion angle during the swing phase of gait in post-stroke individuals

 THIS DOES NOTHING TO GET SURVIVORS RECOVERED! If you can't write EXACT protocols for guaranteed recovery, then get the hell out of stroke! Describing something does nothing for survivors! And you are too blitheringly stupid to see that; along with your mentors and seniors researchers! I'd have you all fired!

Relationship between lower limb muscle coordination and knee flexion angle during the swing phase of gait in post-stroke individuals

• Taiga Karino,
• Toshiyuki Aoyama,
• Kenichi Yoshikawa,
• Kiyoshige Ishibashi &
• Yutaka Kohno 

We are providing an unedited version of this manuscript to give early access to its findings. Before final publication, the manuscript will undergo further editing. Please note there may be errors present which affect the content, and all legal disclaimers apply.

Abstract

Background

Stroke patients with hemiplegia often show inefficient gait patterns, including reduced knee flexion during the swing phase, which may increase fall risk. Post-stroke gait frequently involves merged muscle synergies that affect lower limb kinematics. However, it remains unclear how muscle synergy merging and fractionation relate to knee flexion during the swing phase. Therefore, this study aimed to examine the association between knee flexion during the swing phase and muscle synergy merging and fractionation patterns in patients with stroke.

Methods

The study comprised 21 stroke patients with hemiplegia. Surface electromyography was recorded from eight lower-limb muscles on the paretic side during comfortable gait. Maximum knee flexion angle (MKFA) during the swing phase was measured using a markerless motion capture system. Using non-negative matrix factorization, the number of muscle synergies, their spatiotemporal structure were calculated. Participants were classified into a low-synergy group (LS; n = 5; one or two synergies) or a high-synergy group (HS; n = 16; three synergies). Group comparisons of MKFA during the swing phase were performed. Furthermore, we investigated whether muscle synergies of the HS group could be fractionations of those of the LS group.

Results

The HS group showed significantly greater MKFA compared with the LS group (p = 0.032). In the HS group, the ankle plantar flexors constituted an independent muscle synergy, whereas in the LS group, these muscles had high weightings within a muscle synergy associated with load response. Furthermore, the independent muscle synergies observed in the HS group were shown to be fractionated from the merged muscle synergies present in the LS group.

Conclusion

Our results showed that merged muscle synergies were associated with reduced MKFA during the swing phase, whereas an independent synergy involving the plantar flexors was associated with greater knee flexion. These findings suggest that fractionation of the plantar flexor synergy may be important for improving knee kinematics after stroke and could inform targeted rehabilitation strategies. Given the relatively small and imbalanced sample size, cautious interpretation of the findings is warranted. Further studies with larger, balanced samples are needed to further strengthen the evidence for these findings.

Muscle synergy-driven ensemble learning framework for individualized stroke gait rehabilitation

 Great word salad but is anything in here even approaching a protocol that will deliver EXACT RECOVERY? Since the answer is no; YOU'RE ALL FIRED! The goal of stroke research is survivor recovery! And your mentors and senior researchers are THAT FUCKING INCOMPETENT?

Muscle synergy-driven ensemble learning framework for individualized stroke gait rehabilitation

n

• Jaehyuk Lee & 
• Eunchan Kim 

Scientific Reports volume 15, Article number: 44025 (2025) Cite this article

Abstract

This study proposes a novel ensemble machine learning (ML) framework integrating neurophysiological principles from muscle synergy analysis to support clinical decisions in stroke gait rehabilitation. The framework leverages spatial and temporal features of muscle synergies using a hierarchical ensemble model to improve classification performance and interpretability. Muscle synergies were extracted using non-negative matrix factorization from surface EMG recordings of 380 participants, comprising 120 healthy and 260 post-stroke individuals, each contributing one leg. Feature vectors were derived through a bidirectional decomposition process, wherein either the spatial (W) or temporal (H) synergy matrix was fixed to normative patterns obtained from healthy controls. This enabled the extraction of patient-specific deviations in coordination and activation timing, serving as interpretable indicators of stroke-related neuromuscular impairment. Separate classifiers trained on each feature domain were integrated via meta-regression, achieving classification accuracies above 98% across all configurations on the internal test dataset. After training, performance-weighted feature importance values from tree-based models validated the clinical relevance of learned classification criteria. SHAP (SHapley Additive exPlanations) values quantified sample-specific feature contributions on the test dataset, ensuring individual-level interpretability of predictions. The proposed framework bridges stroke-related neuromuscular impairments and clinical insights, laying a foundation for integrating neurophysiologically grounded ML models into rehabilitation.

Introduction

In clinical settings, precise movement assessment facilitates the early diagnosis, treatment planning, and monitoring of neuromuscular impairments. Consequently, movement analysis has become increasingly crucial, particularly in rehabilitation medicine, as it provides a comprehensive understanding of motor function, optimizes performance, and informs individualized rehabilitation strategies1. Over the past few decades, the development of movement analysis has accelerated, particularly with the advent of muscle synergy analysis (MSA), which employs unsupervised machine learning (ML) techniques to elucidate the neurophysiological control of movement2. A key principle of MSA is to uncover the spatial and temporal coordination of the activation of multiple muscles to produce efficient movement, rather than controlling each muscle independently3. Currently, in the fields of neurological4 and musculoskeletal5 diseases, MSA has provided valuable insights into motor control mechanisms by factorizing coordinated activation into spatial and temporal features.

Among various neuromuscular impairments, stroke has been the most frequently investigated using MSA to identify its gait features, due to its high incidence rate6. Notably, stroke is the second leading cause of death worldwide7 and causes persistent post-survival motor impairments that significantly affect the quality of life8. Many studies have elucidated specific patterns of gait dysfunction in stroke patients, such as the phenomenon of synergy merging, in which multiple muscle synergies combine into a single, less efficient, and specific synergy during gait9,10. Recovery from these pathological gait patterns is therefore considered a critical target in stroke rehabilitation to restore more normalized gait.

Several prior studies have applied muscle synergy features, particularly temporal coefficients, in combination with tree-based ML models to assess motor impairment and predict rehabilitation outcomes in stroke patients. For example, Yang et al. demonstrated the utility of temporal muscle synergy features by applying random forest classifiers to characterize impairment severity during sit-to-stand tasks11 and to estimate the effects of short-term rehabilitation12. An et al. further extended this approach to kinetic assessments using handrail force data in severely affected patients13. While these studies focused on temporal features to assess functional states, our study integrates both gait-specific temporal and spatial features, placing particular emphasis on model-level interpretability for pre-rehabilitation decision-making.

By utilizing MSA along with knowledge to understand factorized features, clinicians and practitioners can develop targeted rehabilitation protocols, optimize training strategies, and refine therapeutic interventions to improve gait performance in stroke patients. However, its clinical adoption remains limited due to ambiguities in interpreting and prioritizing synergy features, which often requires substantial practitioner expertise and introduces variability in application14,15,16. Therefore, in this study, we propose that an ensemble ML framework that highlights clinically relevant synergies may facilitate effective translation into practice. To the best of our knowledge, despite the introduction of various ML models and advancements in computer-aided diagnostic systems designed to reduce subjectivity and improve the consistency in assessments using electromyography (EMG)17,18, the application of MSA for therapeutic decision-making within an ML framework remains extremely rare. In particular, our approach introduces a bidirectional decomposition strategy anchored to normative synergy templates and incorporates SHAP (SHapley Additive exPlanations)-based individualization to enhance interpretability and clinical relevance.

In addition to their use in clinical diagnostics, EMG signals have been widely applied in robotics to train ML algorithms for controlling prosthetic arms and legs. This process typically involves a series of preprocessing steps such as signal filtering, normalization, and feature extraction in both the time and frequency domains. Among these steps, segmentation window size is critical, as it influences both temporal resolution and feature reliability. Shorter windows enhance temporal precision but increase variability, while longer windows stabilize signals but may miss transient events19. Commonly extracted features include root mean square, mean absolute value, median frequency, and wavelet coefficients. These features are then used as inputs for various ML models, including support vector machines, random forests, and deep learning architectures, to predict motor intentions and classify movement patterns. However, these studies have primarily focused on improving the model performance, often overlooking the interpretability of the contributing features or the physiological relevance of the learned representations20,21.

To address these limitations, this study proposed a framework aimed at enhancing the clinical interpretability of EMG data by extracting physiologically meaningful features using MSA. These features were subsequently used to train an ensemble ML model. Feature importance metrics and SHAP analysis were applied to identify synergy components that most influenced the model’s output, enabling individualized prioritization of muscle synergies, which refers to selecting the most relevant features for each patient based on SHAP-derived importance values. Furthermore, a voting-based classification approach was adopted by integrating four distinct tree-based ML models to ensure performance robustness across varying data characteristics. To support reliable interpretation, model-specific weights were assigned based on each model’s predictive performance when aggregating SHAP values and feature importance.

Notably, the spatial and temporal features derived from MSA exhibit distinct structural and statistical characteristics. Separate voting classification models were developed for each feature type and subsequently integrated into a unified ensemble model (Fig. 1). This approach provides a unified framework for decision making by consolidating predictions from both feature sets into a single model.

Accordingly, the proposed framework may offer several stage-specific advantages along the analytical pipeline. First, the use of non-negative matrix factorization (NMF) to extract muscle co-activation patterns (spatial features) and activation timings (temporal features) allows clinicians to identify which muscle groups and gait phases may be most influential in each patient. Second, these quantified features can be translated into a preliminary prioritization scheme, guiding therapists toward muscle synergies that the model indicates as clinically significant. Third, because the framework summarizes its outputs using familiar visualization tools such as feature importance scores and SHAP plots, it supports clinical interpretation without requiring expertise in ML, potentially improving usability in practice.

Fig. 1

Overview of the muscle synergy-driven ML framework for stroke gait rehabilitation: from EMG to SHAP-informed therapeutic decision-making.

Full size image

Results

Extracted muscle synergy

Four muscle synergies were extracted from all participants based on the criterion that collectively accounted for over 90% of the EMG variance in healthy controls (variance accounted for, VAF = 0.95 ± 0.03). This number aligns with previously reported synergy structures identified during human locomotion22,23.

Numerous studies on MSA have demonstrated that NMF serves as a robust method for uncovering modular motor control strategies, as it enforces physiological non-negativity constraints and aligns closely with the structure of real-world muscle activation patterns22,23. By grouping co-activated muscles into low-dimensional modules, NMF captures the functional coordination among individual muscles, thereby enhancing the interpretability of complex motor behaviors. In this study, each extracted synergy was explicitly linked to its physiological role based on dominant muscle contributions and their established biomechanical functions during gait.

As summarized in Table1, Synergy 1 (S1) predominantly involves the hamstring muscles, including the biceps femoris (BFM) and semimembranosus (SEM), functioning primarily for deceleration and stabilization of the leg during the terminal swing and early stance phases. Synergy 2 (S2), characterized by strong activation of Rectus Femoris (RFM) and Adductor Magnus (ADD), contributes to knee extension and hip adduction, playing a key role in load acceptance and medial-lateral control during the early stance phase. Synergy 3 (S3), predominantly involving the Gastrocnemius (GCM), is critical for forward propulsion and push-off during terminal stance and pre-swing phases. Finally, Synergy 4 (S4), involving Gluteus Medius (GLU), Vastus Medialis (VMM), and Tibialis Anterior (TIB), supports pelvic stability and stance phase control, particularly during mid-stance24,25..

As shown in Fig. 2, visual comparison revealed that in patients with stroke, the muscles contributing to individual synergies (spatial features) were generally more strongly activated than in normal controls, and the timing of synergy activation (temporal features) appeared ambiguous and delayed.

Moreover, to assess the impact of stroke on synergy organization, we applied a bidirectional decomposition strategy. Extracting the spatial synergy features of stroke patients with fixed temporal features from healthy controls revealed a significant decrease in VAF (0.19 ± 0.16), indicating severely disrupted muscle coordination post-stroke. Conversely, extracting temporal synergy features with fixed spatial patterns resulted in a relatively higher VAF (0.70 ± 0.10). These findings support the hypothesis that spatial features (muscle coordination) are more substantially affected by stroke compared to temporal features (activation timing), consistent with previous clinical observations of altered neuromuscular recruitment patterns in post-stroke gait25,26.

By clearly relating NMF-derived synergy structures to established biomechanical functions and clinical phenomena, our proposed framework significantly enhances interpretability and facilitates clinical decision-making for targeted and individualized gait rehabilitation.

More at link.

Motor modules are largely unaffected by pathological walking biomechanics: a simulation study

 Ask your competent? doctor how this information is going to recover your walking ability. In my case since most of my motor and pre-motor cortex is dead, nothing here is useful. My dead brain destroyed any possible synergy there was. WHAT IS THE FIX?

Motor modules are largely unaffected by pathological walking biomechanics: a simulation study

• Mohammad Rahimi Goloujeh &
• Jessica L. Allen 

Journal of NeuroEngineering and Rehabilitation volume 22, Article number: 16 (2025) Cite this article

• Metrics details

Abstract

Background

Motor module (a.k.a. muscle synergy) analysis has frequently been used to provide insight into changes in muscle coordination associated with declines in walking performance, to evaluate the effect of different rehabilitation interventions, and more recently, to control exoskeletons and prosthetic devices. However, it remains unclear whether changes in muscle coordination revealed via motor module analysis stem from abnormal walking biomechanics or neural control. This distinction has important implications for the use of motor module analysis for rehabilitation interventions and device design. Thus, this study aims to elucidate the extent to which motor modules emerge from pathological walking biomechanics, i.e. abnormal walking biomechanics commonly observed in individuals with neurological disease and/or injury.

Methods

We conducted a series of computer simulations using OpenSim Moco to simulate pathological walking biomechanics by manipulating speed, asymmetry, and step width in a three-dimensional musculoskeletal model. We focused on these spatiotemporal metrics because they are commonly altered in individuals with Parkinson’s disease, stroke survivors, etc. and have been associated with changes in motor module number and structure. We extracted motor modules using nonnegative matrix factorization from the muscle activations from each simulation. We then examined how alterations in walking biomechanics influenced the number and structure of extracted motor modules and compared the findings to previous experimental studies.

Results

The motor modules identified from our simulations were similar to those identified from previously published experiments of non-pathological walking. Moreover, our findings indicate that the same motor modules can be used to generate a range of pathological-like waking biomechanics by modulating their recruitment over the gait cycle. These results contrast with experimental studies in which pathological-like walking biomechanics are accompanied by a reduction in motor module number and alterations in their structure.

Conclusions

This study highlights that pathological walking biomechanics do not necessarily require abnormal motor modules. In other words, changes in number and structure of motor modules can be a valuable indicator of alterations in neuromuscular control and may therefore be useful for guiding rehabilitation interventions and controlling exoskeletons and prosthetic devices in individuals with impaired walking function due to neurological disease or injury.

Background

Motor module (a.k.a. muscle synergy) analysis has emerged over the last few decades as a useful method to characterize the complex coordination of the numerous muscles involved in movements such as walking. Motor modules reflect coordinated patterns of muscle activity that can be flexibly combined to meet the goals of different movement behaviors [1]. It is hypothesized that motor modules reflect an underlying nervous system strategy to overcome the complexity of controlling movement by grouping muscles into functional units. As such, many researchers are using motor modules to evaluate the effect of different rehabilitation interventions on neuromuscular control [2,3,4,5,6,7,8,9,10,11,12] and even to control exoskeletons and prosthetic devices [13,14,15,16,17,18,19,20,21]. However, since motor modules are identified from experimentally-recorded electromyography (EMG) data using numerical decomposition techniques such as principal component analysis or nonnegative matrix factorization [1, 22, 23], there is ongoing debate regarding whether they truly represent an underlying neural strategy or simply emerge from the biomechanics of the recorded movement. Incorporating EMG-derived motor modules into rehabilitation interventions and/or into controllers for exoskeletons and prosthetics introduces additional complexity compared to recording only movement biomechanics. If motor modules are simply emergent from movement biomechanics, then there may be no need to incorporate such complexity into these settings. Therefore, it is critical to understand to what extent EMG-derived motor modules reflect an underlying neural strategy to support their use in identifying neuromuscular deficits limiting walking function, guiding rehabilitation efforts, and facilitating control of exoskeleton and prosthetic devices.

Converging evidence suggests that recruiting a reduced number of motor modules to control walking contributes to impaired walking function. Several studies have identified that between four to six motor modules are needed to describe muscle activity in unimpaired walking [24,25,26]. Comparatively fewer motor modules are required to describe muscle activity in people with neurological deficits (e.g., stroke [7, 24, 27, 28] cerebral palsy [29] and Parkinson’s disease [30,31,32]) and musculoskeletal conditions such as osteoarthritis [33]. Given that each module in unimpaired walking is organized around producing biomechanical functions such as leg swing control and forward propulsion [34, 35], it is not surprising that individuals with a reduced number of motor modules typically walk with impaired walking function. For example, reduced motor module number in stroke survivors is associated with slower walking speeds and more asymmetrical steps [24, 27] and an inability to change speed, cadence, step length, and step height [28]. Moreover, increases in motor module number that occur with rehabilitation are associated with improved walking function [10, 36]. While the prevailing interpretation of these studies is that a decrease in motor module number causes impaired function, an alternative explanation is that the observed reduction in motor modules may be a consequence of the altered walking biomechanics rather than the cause.

Musculoskeletal modeling and simulation offer a means to disentangle the effects of neural control and biomechanics on motor modules in a carefully controlled manner. For example, Falisse et al., demonstrated a reduction in motor module number alone could not produce the crouch gait biomechanics often observed in cerebral palsy [37]. Moreover, Mehrabi et al., found that normal walking biomechanics could not be achieved with reduced motor module number [38]. Conversely, a recent experimental study provided some evidence that healthy individuals could emulate crouch gait patterns commonly observed in children with cerebral palsy without reducing motor module number [39]. Taken together, these studies suggest that although biomechanics may have some influence on motor modules, there remains room for them to have a neural basis. However, a limitation of these studies is their focus on a single gait type, preventing insights into the extent of biomechanics versus neural control on motor module structure across the diverse walking biomechanics often observed in populations with neurological disease and/or injury.

The purpose of this study was to explore the extent to which motor modules are emergent from pathological walking biomechanics. Or in other words, does pathological walking biomechanics require a reduced number and/or altered structure of motor modules. Inspired by the work of De Groote et al. [26] in which motor modules were extracted from individual muscle-driven simulations of unimpaired walking identical to how motor modules are identified experimentally from EMG, we compared motor modules extracted from 27 muscle-driven simulations of pathological gait biomechanics. We use the term “pathological walking biomechanics” to refer to measurable gait characteristics such as spatiotemporal parameters, kinematics, and kinetics resulting from injury or disease. In this study, we focused on the effect of varying speeds, step length asymmetry, and step widths commonly observed in populations with neurological disease or injury such as stroke survivors, persons with Parkinson’s disease, etc. We hypothesized that motor modules cannot be explained by biomechanics alone and thus reflect to some extent an underlying neural control strategy. Based on this hypothesis, we predicted that the number and structure of motor modules would not differ across simulations with different pathological walking biomechanics.

More at link.

Quantitative Evaluation of Motion Compensation in Post-stroke Rehabilitation Training Based on Muscle Synergy

 If anything here gets survivors recovered it  is well hidden and thus useless for survivors.

Quantitative Evaluation of Motion Compensation in Post-stroke Rehabilitation Training Based on Muscle Synergy

Yanhong Liu¹ Yaowei Li^{1, 2*} Zan Zhang^1* Benyan Huo¹ Anqin Dong³

• ¹School of Electrical and Information Engineering, Zhengzhou University, China
• ²Zhengzhou University, China
• ³The Rehabilitation Department, Fifth Affiliated Hospital of Zhengzhou University, China

The final, formatted version of the article will be published soon.

Stroke is the second leading cause of death globally and a primary factor contributing to disability. Unilateral limb motor impairment caused by stroke is the most common scenario. The bilateral movement pattern plays a crucial role in assisting stroke survivors on the affected side to relearn lost skills. However, motion compensation often lead to decreased coordination between the limbs on both sides. Furthermore, muscle fatigue resulting from imbalanced force exertion on both sides of the limbs can also impact the rehabilitation outcomes.In this study, an assessment method based on muscle synergy indicators was proposed to objectively quantify the impact of motion compensation issues on rehabilitation outcomes.Muscle synergy describes the body's neuromuscular control mechanism, representing the coordinated activation of multiple muscles during movement. 8 post-stroke hemiplegia patients and 8 healthy subjects participated in this study. During hand-cycling tasks with different resistance levels, surface electromyography signals were synchronously collected from these participants before and after fatigue. Additionally, a simulated compensation experiment was set up for healthy participants to mimic various hemiparetic states observed in patients.: Synergy symmetry and synergy fusion were chosen as potential indicators for assessing motion compensation. The experimental results indicate significant differences in synergy symmetry and fusion levels between the healthy control group and the patient group (p ≤ 0.05), as well as between the healthy control group and the compensation group. Moreover, the analysis across different resistance levels showed no significant variations in the assessed indicators (p > 0.05), suggesting the utility of synergy symmetry and fusion indicators for the quantitative evaluation of compensation behaviors. Although muscle fatigue did not significantly alter the symmetry and fusion levels of bilateral synergies (p > 0.05), it did reduce the synergy repeatability across adjacent movement cycles, compromising movement stability and hindering patient recovery. Based on synergy symmetry and fusion indicators, the degree of bilateral motion compensation in patients can be quantitatively assessed, providing personalized recommendations for rehabilitation training and enhancing its effectiveness.

Keywords: Rehabilitation training, motion compensation, surface electromyography, Muscle Synergy, Quantitative assessment

Received: 23 Jan 2024; Accepted: 20 Feb 2024.

Copyright: © 2024 Liu, Li, Zhang, Huo and Dong. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Quantification of Upper-Limb Motor Function for Stroke Rehabilitation Through Manifold Similarity of Muscle Synergy

 I have no clue. Ask your competent? doctor how this is going to get you 100% recovered. I consider assessments worthless unless they point directly to the EXACT STROKE REHAB PROTOCOLS that get you recovered!

Quantification of Upper-Limb Motor Function for Stroke Rehabilitation Through Manifold Similarity of Muscle Synergy

Xiao Li; Hong Zeng; Yongqiang Li; Aiguo Song

All Authors

• 
  
• 
  

• 
  
• 
  
• 
  
• 
  

---

Abstract

Authors

Keywords

Abstract:

Quantifying post-stroke patient motor function is important for assessing rehabilitation progress and optimizing the behavior of adaptive rehabilitation robots. To this end, researchers have increasing turned to the concept of muscle synergies, which encodes the simplified neuromuscular control strategy employed by the central nervous system in response to post-stroke impairment. In essence, the assessment metrics should possess two key attributes: the ability to differentiate between individuals in the pathological and healthy groups, and the capacity to yield consistent measurements within the same individual, thereby facilitating the refinement of adaptive control algorithms. Recent findings have indicated that employing manifold similarity measurements can enhance the class separability and intra-class compactness for the classification/clustering algorithm. Consequently, we hypothesize that evaluating synergy and synergy activation similarities, while considering the underlying manifold structure, will render a more sensitive and reliable approach for quantifying motor function in post-stroke patients. To validate our hypothesis, we conducted a study involving twenty healthy subjects and ten post-stroke patients. Our results demonstrate that the utilization of manifold similarities leads to superior outcomes compared to conventional metrics based on muscle synergy. Specifically, we observed higher sensitivity ( gw v.s. Sw , 0.0457 v.s. 0.0030 ), greater intra-subject reliability ( gc v.s. Sc , 0.6060 v.s. 0.1081 ), and stronger correlations with clinical scores ( gw v.s. Sw , 0.7588 v.s. 0.6249 ) than conventional metrics. Therefore, the proposed similarity metrics may be promising for transferring to adaptive control of rehabilitation robots.

Published in: IEEE Robotics and Automation Letters ( Early Access )