Advancing Lower-Limb Post-Stroke Rehabilitation: Leveraging Muscle Synergies and Multichannel Functional Electrical Stimulations
基於肌肉協同和多通道功能電刺激的中風病人下肢康復研究
Student thesis: Doctoral Thesis
Author(s)
Related Research Unit(s)
Detail(s)
Awarding Institution | |
---|---|
Supervisors/Advisors |
|
Award date | 19 Sept 2023 |
Link(s)
Permanent Link | https://scholars.cityu.edu.hk/en/theses/theses(832cf85a-c466-497e-b170-2caeb71a3e98).html |
---|---|
Other link(s) | Links |
Abstract
In the field of neural mechanisms underlying motor coordination in the central nervous system (CNS), muscle synergy has emerged as a leading explanation for motor control organization. The CNS accomplishes complex motor tasks by activating multiple muscles in unison rather than individually. Numerous studies have indirectly supported the muscle synergy hypothesis, and some research has even considered muscle synergies as physiological markers of motor cortical damage in conditions such as stroke or trauma. These findings have informed the development of rehabilitation strategies and investigations into motor control recovery. This thesis investigates muscle synergies across age groups and employs a factorization algorithm to identify inhibitory components, aiming to uncover the underlying neural mechanisms of motor control during human overground walking. Furthermore, the study proposes an innovative and efficient functional electrical stimulation (FES) approach for lower-limb post-stroke rehabilitation.
As individuals age, the neuromuscular control of overground walking undergoes modifications. While prior research has established that gait biomechanics change with age, it remains uncertain whether these alterations are accompanied by concurrent age-related modulations in neuromuscular control. Therefore, in our first part, we analyzed gait kinematics and electromyographic signals (EMGs; 14 lower-limb and trunk muscles) collected at three speeds during overground walking in 11 healthy young adults (N=11; 7 females; 23.4 \pm 2.5 years) and 11 healthy elderlies (N=11; 7 females; 67.2 \pm 4.3 years). Neuromuscular control was characterized by extracting muscle synergies from EMGs and the synergies of both groups were \emph{k}-means-clustered. The synergies of the two groups were grossly similar, but we observed numerous cluster- and muscle-specific differences between the age groups. At the population level, some hip-motion-related synergy clusters were more frequently identified the elderly, while others were more frequent in young adults. Such differences in synergy prevalence between the age groups are consistent with the finding that elderlies had a larger hip flexion range. For the synergies shared between both groups, the elderlies had higher inter-subject variability of the temporal activations than young adults. To further explore what synergy characteristics may be related to this inter-subject variability, we found that the inter-subject variance of temporal activations correlated negatively with the sparseness of the synergies in elderlies but not young adults during slow walking. Overall, our results suggest that as humans age, not only are the muscle synergies for walking fine-tuned in structure, but their temporal activation patterns are also more heterogeneous across individuals, possibly reflecting individual differences in prior sensorimotor experience or ageing-related changes in limb neuro-musculoskeletal properties.
Non-negative matrix factorization (NMF), widely used in motor neuroscience for identifying muscle synergies from electromyographical signals (EMGs), extracts non-negative synergies and is yet unable to identify potential negative components in synergies underpinned by inhibitory spinal interneurons. To overcome this constraint, in our second part, we propose to utilize the rectified latent variable model (RLVM) to extract muscle synergies. RLVM uses an autoencoder neural network, and the weight matrix of its neural network could be negative, while latent variables must remain non-negative. If inputs to the model are EMGs, the weight matrix and latent variables are muscle synergies and their temporal activation coefficients, respectively. We compare the performances of NMF and RLVM in identifying muscle synergies in simulated and experimental datasets. Our simulation results showed that RLVM performed better in identifying muscle-synergy subspace, but NMF outputs had a higher correlation with ground truth even corrupted with signal-dependent or Gaussian noise. Finally, we applied RLVM to a previously published experimental dataset comprising EMGs from upper-limb muscles and spike recordings of spinal premotor interneurons during voluntary grasping of two monkeys. RLVM and NMF synergies were highly similar, but a few minor negative muscle components were observed in RLVM synergies. The muscles with negative components identified by RLVM were close to 0 in their corresponding synergies identified by NMF. Also, the negative components of RLVM synergies agreed with the muscle connectivity of the premotor interneurons with inhibitory muscle fields, as identified by spike-triggered averaging of EMGs. Our results demonstrate the feasibility of RLVM in extracting potential inhibitory muscle-synergy components from EMGs.
Moreover, FES is a promising technology to communicate with the neuromuscular system by sensorimotor system. In the third part, we proposed an innovative FES strategy based on the muscle synergy for lower-limb of post-stroke rehabilitation. A protocol with repeated overground walking assisted by synergy-based FES was conducted in 22 patients (N = 22; 12 females; age 58.3 \pm 7.8 years) with post-stroke hemiparesis. The customized FES was applied to 7 lower-extremity muscles of the patients during walking. The Lite-Gait was applied to patients during walking to avoid their accidentally falling. Envelopes of stimuli were individualized by re-composing the muscle synergies extracted from the healthy subject selected using the cross-fit algorithm. After a five-session with over 600 gait cycles each time, synergy-based FES induced higher increases in Fugl-Meyer scores after the FES. The synergy similarity of the selected synergies to design FES strategy increased significantly than that of non-selected synergies in patients without physical therapy group. In addition, significant improvement was observed in hip flexion/extension, knee flexion and ankle dorsiflexion/plantarflexion in some patients after FES. More importantly, after this short-term FES, statistical differences between affected and unaffected sides disappear in stance ratio and swing time in patients. Ultimately, our results indicate that synergy-based FES therapy induced improvements in post-stroke motor performance, which may contribute to the modifications of muscle synergies.
As individuals age, the neuromuscular control of overground walking undergoes modifications. While prior research has established that gait biomechanics change with age, it remains uncertain whether these alterations are accompanied by concurrent age-related modulations in neuromuscular control. Therefore, in our first part, we analyzed gait kinematics and electromyographic signals (EMGs; 14 lower-limb and trunk muscles) collected at three speeds during overground walking in 11 healthy young adults (N=11; 7 females; 23.4 \pm 2.5 years) and 11 healthy elderlies (N=11; 7 females; 67.2 \pm 4.3 years). Neuromuscular control was characterized by extracting muscle synergies from EMGs and the synergies of both groups were \emph{k}-means-clustered. The synergies of the two groups were grossly similar, but we observed numerous cluster- and muscle-specific differences between the age groups. At the population level, some hip-motion-related synergy clusters were more frequently identified the elderly, while others were more frequent in young adults. Such differences in synergy prevalence between the age groups are consistent with the finding that elderlies had a larger hip flexion range. For the synergies shared between both groups, the elderlies had higher inter-subject variability of the temporal activations than young adults. To further explore what synergy characteristics may be related to this inter-subject variability, we found that the inter-subject variance of temporal activations correlated negatively with the sparseness of the synergies in elderlies but not young adults during slow walking. Overall, our results suggest that as humans age, not only are the muscle synergies for walking fine-tuned in structure, but their temporal activation patterns are also more heterogeneous across individuals, possibly reflecting individual differences in prior sensorimotor experience or ageing-related changes in limb neuro-musculoskeletal properties.
Non-negative matrix factorization (NMF), widely used in motor neuroscience for identifying muscle synergies from electromyographical signals (EMGs), extracts non-negative synergies and is yet unable to identify potential negative components in synergies underpinned by inhibitory spinal interneurons. To overcome this constraint, in our second part, we propose to utilize the rectified latent variable model (RLVM) to extract muscle synergies. RLVM uses an autoencoder neural network, and the weight matrix of its neural network could be negative, while latent variables must remain non-negative. If inputs to the model are EMGs, the weight matrix and latent variables are muscle synergies and their temporal activation coefficients, respectively. We compare the performances of NMF and RLVM in identifying muscle synergies in simulated and experimental datasets. Our simulation results showed that RLVM performed better in identifying muscle-synergy subspace, but NMF outputs had a higher correlation with ground truth even corrupted with signal-dependent or Gaussian noise. Finally, we applied RLVM to a previously published experimental dataset comprising EMGs from upper-limb muscles and spike recordings of spinal premotor interneurons during voluntary grasping of two monkeys. RLVM and NMF synergies were highly similar, but a few minor negative muscle components were observed in RLVM synergies. The muscles with negative components identified by RLVM were close to 0 in their corresponding synergies identified by NMF. Also, the negative components of RLVM synergies agreed with the muscle connectivity of the premotor interneurons with inhibitory muscle fields, as identified by spike-triggered averaging of EMGs. Our results demonstrate the feasibility of RLVM in extracting potential inhibitory muscle-synergy components from EMGs.
Moreover, FES is a promising technology to communicate with the neuromuscular system by sensorimotor system. In the third part, we proposed an innovative FES strategy based on the muscle synergy for lower-limb of post-stroke rehabilitation. A protocol with repeated overground walking assisted by synergy-based FES was conducted in 22 patients (N = 22; 12 females; age 58.3 \pm 7.8 years) with post-stroke hemiparesis. The customized FES was applied to 7 lower-extremity muscles of the patients during walking. The Lite-Gait was applied to patients during walking to avoid their accidentally falling. Envelopes of stimuli were individualized by re-composing the muscle synergies extracted from the healthy subject selected using the cross-fit algorithm. After a five-session with over 600 gait cycles each time, synergy-based FES induced higher increases in Fugl-Meyer scores after the FES. The synergy similarity of the selected synergies to design FES strategy increased significantly than that of non-selected synergies in patients without physical therapy group. In addition, significant improvement was observed in hip flexion/extension, knee flexion and ankle dorsiflexion/plantarflexion in some patients after FES. More importantly, after this short-term FES, statistical differences between affected and unaffected sides disappear in stance ratio and swing time in patients. Ultimately, our results indicate that synergy-based FES therapy induced improvements in post-stroke motor performance, which may contribute to the modifications of muscle synergies.