Impact of Rounded Back Posture on Motor Unit Potentials and Fascicle Length of Shoulder Retractors in Children.

OBJECTIVE: There is little proof to determine the features of the muscles' motor unit potentials (MUPs) in children with poor posture. Current evaluation could be of value for future studies as a reference. The purpose was to detect the impact of rounded back posture on the characteristics of the MUPs and fascicle length of the shoulder retractors in children. METHODS: Participants in this study were 60 children (boys and girls), their ages were from 7 to 10 years old. Children were allocated into healthy children group (A) and rounded back posture group (B). MUPs and fascicle length of middle trapezius were assessed by electromyography and ultrasonography respectively. RESULTS: When compared to the normal group, the rounded back group's right and left middle trapezius MUPs count and amplitude significantly increased. As regards to the middle trapezius MUPs duration between the two groups, there was no significant difference. Also, the rounded back posture group exhibited significantly lower fascicle length in middle trapezius of both sides than the normal group. CONCLUSION: Forward shoulder posture is accompanied by atypical middle trapezius MUPs characteristics and also lowered fascicle length. Thus, children with forward-leaning posture could increase the likelihood of developing any of the many shoulder disorders.

Assessment of inter-rater and intra-rater reliability of the Luna EMG robot as a tool for assessing upper limb proprioception in patients with stroke-a prospective observational study.

Background: The aim of the study was to assess the inter-rater and intra-rater agreement of measurements performed with the Luna EMG (electromyography) multifunctional robot, a tool for evaluation of upper limb proprioception in individuals with stroke. Methods: The study was conducted in a group of patients with chronic stroke. A total of 126 patients participated in the study, including 78 women and 48 men, on average aged nearly 60 years (mean = 59.9). Proprioception measurements were performed using the Luna EMG diagnostic and rehabilitation robot to assess the left and right upper limbs. The examinations were conducted by two raters, twice, two weeks apart. The results were compared between the raters and the examinations. Results: High consistency of the measurements performed for the right and the left hand was reflected by the interclass correlation coefficients (0.996-0.998 and 0.994-0.999, respectively) and by Pearson's linear correlation which was very high (r = 1.00) in all the cases for the right and the left hand in both the inter-rater and intra-rater agreement analyses. Conclusions: Measurements performed by the Luna EMG diagnostic and rehabilitation robot demonstrate high inter-rater and intra-rater agreement in the assessment of upper limb proprioception in patients with chronic stroke. The findings show that Luna EMG is a reliable tool enabling effective evaluation of upper limb proprioception post-stroke.

From suspicion to diagnosis: exploration strategy for suspected amyotrophic lateral sclerosis.

The diagnosis of amyotrophic lateral sclerosis (ALS) is based on evidence of upper and lower motor neuron degeneration in the bulbar, cervical, thoracic, and lumbar regions in a patient with progressive motor weakness, in the absence of differential diagnosis. Despite these well-defined criteria, ALS can be difficult to diagnose, given the wide variety of clinical phenotypes. Indeed, the central or peripheral location of the disease varies with a spectrum ranging from predominantly central to exclusively peripheral, symptoms can be extensive or limited to the limbs, bulbar area or respiratory muscles, and the duration of the disease may range from a few months to several decades. In the absence of a specific test, the diagnostic strategy relies on clinical, electrophysiological, biological and radiological investigations to confirm the disease and exclude ALS mimics. The main challenge is to establish a diagnosis based on robust clinical and paraclinical evidence without delaying treatment initiation by increasing the number of additional tests. This approach requires a thorough knowledge of the phenotypes of ALS and its main differential diagnoses.

The diagnosis of amyotrophic lateral sclerosis (ALS) is based on progressive degeneration of upper and lower motor neurons.ALS can be difficult to diagnose due to the wide range of clinical phenotypes (central/peripheral location, symptom distribution, disease duration).A thorough diagnostic strategy including clinical, electrophysiological, biological and radiological investigations is essential to confirm ALS and exclude differential diagnoses.

Acquisition and Analysis of Facial Electromyographic Signals for Emotion Recognition.

The objective of the article is to recognize users' emotions by classifying facial electromyographic (EMG) signals. A biomedical signal amplifier, equipped with eight active electrodes positioned in accordance with the Facial Action Coding System, was used to record the EMG signals. These signals were registered during a procedure where users acted out various emotions: joy, sadness, surprise, disgust, anger, fear, and neutral. Recordings were made for 16 users. The mean power of the EMG signals formed the feature set. We utilized these features to train and evaluate various classifiers. In the subject-dependent model, the average classification accuracies were 96.3% for KNN, 94.9% for SVM with a linear kernel, 94.6% for SVM with a cubic kernel, and 93.8% for LDA. In the subject-independent model, the classification results varied depending on the tested user, ranging from 91.4% to 48.6% for the KNN classifier, with an average accuracy of 67.5%. The SVM with a cubic kernel performed slightly worse, achieving an average accuracy of 59.1%, followed by the SVM with a linear kernel at 53.9%, and the LDA classifier at 41.2%. Additionally, the study identified the most effective electrodes for distinguishing between pairs of emotions.

Unravelling Influence Factors in Pattern Recognition Myoelectric Control Systems: The Impact of Limb Positions and Electrode Shifts.

Pattern recognition (PR)-based myoelectric control systems can naturally provide multifunctional and intuitive control of upper limb prostheses and restore lost limb function, but understanding their robustness remains an open scientific question. This study investigates how limb positions and electrode shifts-two factors that have been suggested to cause classification deterioration-affect classifiers' performance by quantifying changes in the class distribution using each factor as a class and computing the repeatability and modified separability indices. Ten intact-limb participants took part in the study. Linear discriminant analysis (LDA) was used as the classifier. The results confirmed previous studies that limb positions and electrode shifts deteriorate classification performance (14-21% decrease) with no difference between factors (p > 0.05). When considering limb positions and electrode shifts as classes, we could classify them with an accuracy of 96.13 ± 1.44% and 65.40 ± 8.23% for single and all motions, respectively. Testing on five amputees corroborated the above findings. We have demonstrated that each factor introduces changes in the feature space that are statistically new class instances. Thus, the feature space contains two statistically classifiable clusters when the same motion is collected in two different limb positions or electrode shifts. Our results are a step forward in understanding PR schemes' challenges for myoelectric control of prostheses and further validation needs be conducted on more amputee-related datasets.

Muscle Network Connectivity Study in Diabetic Peripheral Neuropathy Patients.

Diabetic peripheral neuropathy (DPN) is a prevalent complication of chronic diabetes mellitus and has a significant impact on quality of life. DPN typically manifests itself as a symmetrical, length-dependent sensorimotor polyneuropathy with severe effects on gait. Surface electromyography (sEMG) is a valuable low-cost tool for assessing muscle activation patterns and precise identification of abnormalities. For the present study, we used information theory methods, such as cross-correlation (CC), normalized mutual information (NMI), conditional granger causality (CG-Causality), and transfer entropy (TE), to evaluate muscle network connectivity in three population groups: 33 controls (healthy volunteers, CT), 10 diabetic patients with a low risk of DPN (LW), and 17 moderate/high risk patients (MH). The results obtained indicated significant alterations in the intermuscular coupling mechanisms due to diabetes and DPN, with the TE group showing the best performance in detecting differences. The data revealed a significant increase in information transfer and muscle connectivity in the LW group over the CT group, while the MH group obtained significantly lower values for these metrics than the other two groups. These findings highlight the sEMG coupling metrics' potential to reveal neuromuscular mechanisms that could aid the development of targeted rehabilitation strategies and help monitor DPN patients.

Transforming the Anthropomorphic Passive Free-Flow Foot Prosthesis Into a Powered Foot Prosthesis With Intuitive Control and Sensation (Bionic FFF).

INTRODUCTION: Approximately 89% of all service members with amputations do not return to duty. Restoring intuitive neural control with somatosensory sensation is a key to improving the safety and efficacy of prosthetic locomotion. However, natural somatosensory feedback from lower-limb prostheses has not yet been incorporated into any commercial prostheses. MATERIALS AND METHODS: We developed a neuroprosthesis with intuitive bidirectional control and somatosensation and evoking phase-dependent locomotor reflexes, we aspire to significantly improve the prosthetic rehabilitation and long-term functional outcomes of U.S. amputees. We implanted the skin and bone integrated pylon with peripheral neural interface pylon into the cat distal tibia, electromyographic electrodes into the residual gastrocnemius muscle, and nerve cuff electrodes on the distal tibial and sciatic nerves. Results. The bidirectional neural interface that was developed was integrated into the existing passive Free-Flow Foot and Ankle prosthesis, WillowWood, Mount Sterling, OH. The Free-Flow Foot was chosen because it had the highest Index of Anthropomorphicity among lower-limb prostheses and was the first anthropomorphic prosthesis brought to market. Conclusion. The cats walked on a treadmill with no cutaneous feedback from the foot in the control condition and with their residual distal tibial nerve stimulated during the stance phase of walking.

Achieving maximal voluntary contraction of paraspinal muscles requires two tasks: Insight from an EMG study of females with and without adolescent idiopathic scoliosis.

An accurate estimation of maximal voluntary muscle activation is critical for normalisation in scientific studies. Only a handful of studies appropriately normalise muscle activation data when investigating paraspinal muscle activity in populations such as adolescent idiopathic scoliosis (AIS). This neglect compromises the ability to interpret data. The aim of this study was to determine the type of trunk extension task that reliably achieves peak paraspinal muscle activation in participants with and without AIS. Adolescent females with typically developing spines (controls: n = 20, mean[SD] age 13.1[1.8]years), or primary right thoracic AIS (n = 24, age: 13.8[1.5]years, Cobb angle thoracic: 39.5[16.4]°, lumbar: 28.0[11.6]°) performed a series of 3x unresisted and 3x resisted maximal voluntary trunk extensions in prone. Paraspinal muscle activation was recorded bilaterally at two thoracic levels and one lumbar level using surface electromyography (EMG). Muscle activation was highly repeatable within task [ICC 0.77-0.95, all p < 0.01]. At group level, there were no differences in peak muscle activation between tasks irrespective of side (left/right) or vertebral level (Estimate 0.98, 95%CI 0.36 to 2.65, p=0.97). Peak activation was achieved with the unresisted task in 40.5%, and resisted task in 59.5% of the total outcomes (6 recording locations, 44 participants). Individual participant maximum amplitude varied up to 64% (mean[SD]:18[13]%) between the unresisted and resisted tasks. We recommend that both the resisted and unresisted trunk extension tasks are used to increase confidence that a maximum voluntary activation of paraspinal muscles is achieved. Failure to do so could introduce large error in the estimations of muscle activation.

Sex-related differences in corticospinal excitability outcome measures of the biceps brachii during a submaximal elbow flexor contraction.

The purpose of this study was to investigate the effects of sex, muscle thickness, and subcutaneous fat thickness (SFT) on corticospinal excitability outcome measures of the biceps brachii. Eighteen participants (10 males and 8 females) completed this study. Ultrasound was used to assess biceps brachii muscle thickness and the overlying SFT. Transcranial magnetic stimulation (TMS) was used to determine corticospinal excitability by inducing motor-evoked potentials (MEPs) at eight different TMS intensities from 90% to 160% of active motor threshold (AMT) from the biceps brachii during an isometric contraction of the elbow flexors at 10% of maximum voluntary contraction (MVC). Biceps brachii maximal compound muscle action potential (Mmax) was also recorded prior to and after TMS. Males had higher (p < 0.001) biceps brachii muscle thickness and lower SFT, produced higher levels of MVC force and had, on average, higher (p < 0.001) MEP amplitudes at lower (p < 0.05) percentages of maximal stimulator output than females during the 10% elbow flexion MVC. Multiple linear regression modeling revealed that sex was not associated with any of the neurophysiological parameters examined, while SFT showed a positive association with the stimulation intensity required at AMT (p = 0.035) and a negative association with biceps brachii pre-stimulus electromyography (EMG) activity (p = 0.021). Additionally, there was a small positive association between muscle thickness and biceps brachii pre-stimulus EMG activity (p = 0.049). Overall, this study suggests that some measures of corticospinal excitability may be different between the sexes and influenced by SFT and muscle thickness.

Utility of the Repetitive Nerve Stimulation Test and Needle EMG in the Trapezius Muscle for the Early Diagnosis of ALS.

OBJECTIVES: To document the utility of decremental responses in the repetitive nerve stimulation test (RNS) and spontaneous activities in needle electromyography (EMG) in the trapezius muscle for the diagnosis of amyotrophic lateral sclerosis. METHODS: Subjects were retrospectively identified from our EMG database. Cervical spondylosis was represented as a disease control group. We investigated the sensitivity and specificity of RNS and EMG in the trapezius muscle and those of diagnostic criteria including the Gold Coast criteria (GCC). RESULTS: We reviewed 120 patients with amyotrophic lateral sclerosis and 17 patients with cervical spondylosis. "RNS or EMG" achieved the highest sensitivity (85%). The specificity was the highest for RNS (94%). Addition of RNS of the deltoid muscle achieved 98% sensitivity in the upper-limb onset amyotrophic lateral sclerosis. The sensitivity of the GCC was very high (88%). CONCLUSIONS: Neurophysiological parameters investigated in this study having close to 100% specificities or sensitivities are useful as complements to the GCC.

Wireless high-resolution surface facial electromyography mask for discrimination of standardized facial expressions in healthy adults.

Wired high resolution surface electromyography (sEMG) using gelled electrodes is a standard method for psycho-physiological, neurological and medical research. Despite its widespread use electrode placement is elaborative, time-consuming, and the overall experimental setting is prone to mechanical artifacts and thus offers little flexibility. Wireless and easy-to-apply technologies would facilitate more accessible examination in a realistic setting. To address this, a novel smart skin technology consisting of wireless dry 16-electrodes was tested. The soft electrode arrays were attached to the right hemiface of 37 healthy adult participants (60% female; 20 to 57 years). The participants performed three runs of a standard set of different facial expression exercises. Linear mixed-effects models utilizing the sEMG amplitudes as outcome measure were used to evaluate differences between the facial movement tasks and runs (separately for every task). The smart electrodes showed specific activation patterns for each of the exercises. 82% of the exercises could be differentiated from each other with very high precision when using the average muscle action of all electrodes. The effects were consistent during the 3 runs. Thus, it appears that wireless high-resolution sEMG analysis with smart skin technology successfully discriminates standard facial expressions in research and clinical settings.

Machine learning approaches to predict whether MEPs can be elicited via TMS.

BACKGROUND: Transcranial magnetic stimulation (TMS) is a valuable technique for assessing the function of the motor cortex and cortico-muscular pathways. TMS activates the motoneurons in the cortex, which after transmission along cortico-muscular pathways can be measured as motor-evoked potentials (MEPs). The position and orientation of the TMS coil and the intensity used to deliver a TMS pulse are considered central TMS setup parameters influencing the presence/absence of MEPs. NEW METHOD: We sought to predict the presence of MEPs from TMS setup parameters using machine learning. We trained different machine learners using either within-subject or between-subject designs. RESULTS: We obtained prediction accuracies of on average 77 % and 65 % with maxima up to up to 90 % and 72 % within and between subjects, respectively. Across the board, a bagging ensemble appeared to be the most suitable approach to predict the presence of MEPs. CONCLUSIONS: Although within a subject the prediction of MEPs via TMS setup parameter-based machine learning might be feasible, the limited accuracy between subjects suggests that the transfer of this approach to experimental or clinical research comes with significant challenges.

Integration of Motor Unit Filters for Enhanced Surface Electromyogram Decomposition During Varying Force Isometric Contraction.

Muscles generate varying levels of force by recruiting different numbers of motor units (MUs), and as the force increases, the number of recruited MUs gradually rises. However, current decoding methods encounter difficulties in maintaining a stable and consistent growth trend in MU numbers with increasing force. In some instances, an unexpected reduction in the number of MUs can even be observed as force intensifies. To address this issue, in this study, we propose an enhanced decoding method that adaptively reutilizes MU filters. Specifically, in addition to the normal decoding process, we introduced an additional procedure where MU filters are reused to initialize the algorithm. The MU filters are iterated and adapted to the new signals, aiming to decode motor units that were actually activated but cannot be identified due to heavy superimposition. We tested our method on both simulated and experimental surface electromyogram (sEMG) signals. We simulated isometric signals (10%-70%) with known MU firing patterns using experimentally recorded MU action potentials from forearm muscles and compared the decomposition results to two baseline approaches: convolution kernel compensation (CKC) and fast independent component analysis (fastICA). Our method increased the decoded MU number by a rate of 135.4% ± 62.5 % and 63.6% ± 20.2 % for CKC and fastICA, respectively, across different signal-to-noise ratios. The sensitivity and precision for MUs decomposed using the enhanced method remained at the same accuracy level (p <0.001) as those of normally decoded MUs. For the experimental signals, eight healthy subjects performed hand movements at five different force levels (10%-90%), during which sEMG signals were recorded and decomposed. The results indicate that the enhanced process increased the number of decoded MUs by 21.8% ± 10.9 % across all subjects. We discussed the possibility of fully capturing all activated motor units by appropriately reusing previously decoded MU filters and improving the balance of activated motor unit numbers across varying excitation levels.

Effects of hand dominance on myoelectric signal of non­fatigued lumbar multifidus muscle during single arm lifts.

Some evidence indicates that lower back muscles located at the non­dominant side of the body are more fatigue resistant than their opposite counterparts presumably due to preferential use of the dominant hand. The aim of the study was to determine if any distinction exists in the surface electromyographic activity of corresponding contralateral non­fatigued lumbar multifidus (LM) muscles as a function of hand dominance. The relative to maximum root mean square, the median frequency (MdF) and spike shape parameters were computed from the surface myoelectric signals of ipsilateral and contralateral lumbar multifidus muscle of 46 adult healthy subjects (27 right­handed, 19 left­handed) during voluntary contractions evoked by the single arm lifts in prone position. Activation of LM as a contralateral muscle to lifted arm was greater than as ipsilateral muscle, independently of handedness. Regardless if LM performed ipsi­ or contralateral action to the lifted arm, the mean spike amplitude, slope, number of peaks per spike and spike duration were greater and mean spike frequency as well as MdF were smaller in the muscle of dominant than non­dominant side. Combined changes of spike shape measures indicate increased recruitment, lower firing rates and higher synchronization of motor units in the LM of dominant side as compared to its counterpart.

Rectus femoris electromyography signal clustering: Data-driven management of crouch gait in patients with cerebral palsy.

This study aimed to investigate how electromyography (EMG) cluster analysis of the rectus femoris (RF) could help to better interpret gait analysis in patients with cerebral palsy (CP). The retrospective gait data of CP patients were categorized into two groups: initial examination (E1, 881 patients) and subsequent examination (E2, 377 patients). Envelope-formatted EMG data of RF were collected. Using PCA and a combined PSO-K-means algorithm, main clusters were identified. Patients were further classified into crouch, jump, recurvatum, stiff and mild gait for detailed analysis. The clusters (labels) were characterized by a significant peak EMG activity during mid-swing (L1), prolonged EMG activity during stance (L2), and a peak EMG activity during loading response (L3). Notably, L2 contained 76% and 92% of all crouch patients at E1 and E2, respectively. Comparing patients with a crouch gait pattern in L2-E1 and L2-E2, two subgroups emerged: patients with persistent crouch (G1) and patients showing improvement at E2 (G2). The minimum activity of RF during 20-45% of the gait was significantly higher (p = 0.025) in G1 than in G2. A greater chance of improvement from crouch gait might be associated with lower RF activity during the stance phase. Using our findings, we could potentially establish an approach to improve clinical decision-making regarding treatment of patients with CP.

Digital Biomarker for Muscle Function Assessment Using Surface Electromyography With Electrical Stimulation and a Non-Invasive Wearable Device.

Sarcopenia is a comprehensive degenerative disease with the progressive loss of skeletal muscle mass with age, accompanied by the loss of muscle strength and muscle dysfunction. Individuals with unmanaged sarcopenia may experience adverse outcomes. Periodically monitoring muscle function to detect muscle degeneration caused by sarcopenia and treating degenerated muscles is essential. We proposed a digital biomarker measurement technique using surface electromyography (sEMG) with electrical stimulation and wearable device to conveniently monitor muscle function at home. When motor neurons and muscle fibers are electrically stimulated, stimulated muscle contraction signals (SMCSs) can be obtained using an sEMG sensor. As motor neuron activation is important for muscle contraction and strength, their action potentials for electrical stimulation represent the muscle function. Thus, the SMCSs are closely related to muscle function, presumptively. Using the SMCSs data, a feature vector concatenating spectrogram-based features and deep learning features extracted from a convolutional neural network model using continuous wavelet transform images was used as the input to train a regression model for measuring the digital biomarker. To verify muscle function measurement technique, we recruited 98 healthy participants aged 20-60 years including 48 [49%] men who volunteered for this study. The Pearson correlation coefficient between the label and model estimates was 0.89, suggesting that the proposed model can robustly estimate the label using SMCSs, with mean error and standard deviation of -0.06 and 0.68, respectively. In conclusion, measuring muscle function using the proposed system that involves SMCSs is feasible.

Tongue brushing enhances the myoelectric activity of the suprahyoid muscles in older adults: a six-week randomized controlled trial.

Tongue brushing improves respiratory function in older adults. Considering connection between the respiratory-related and suprahyoid muscles, this study aimed to investigate whether tongue-brushing interventions can improve myoelectric activity during respiration. A six-week randomized controlled trial was conducted in Kitakyushu, Japan, with 50 participants aged ≥ 65 years. The participants were allocated to the intervention (tongue brushing with routine oral hygiene) or control (routine oral hygiene alone) groups. Surface electromyography (sEMG) was used to assess the myoelectric activity of the suprahyoid muscles during inhalation, exhalation, and forced vital capacity (FVC). A survey was conducted at baseline and the end of the follow-up period. Thirty-six participants were recruited for the analysis. The root mean squares (RMS) of sEMG during exhalation increased significantly at the end of the follow-up period compared with that at baseline in the intervention group [48.7 (18.0-177.5) vs. 64.9 (21.6-163.0), p = 0.001], but not in the control group. The generalized linear model revealed that the ratio of change in FVC was correlated with the change in the RMS of sEMG of the suprahyoid muscles during exhalation after adjusting for potential confounders. Tongue brushing enhances the myoelectric activity of the suprahyoid muscle.

Diaphragmatic myoclonus post cerebellar hemorrhage: a case report.

BACKGROUND: Diaphragmatic myoclonus is a rare motor disorder that affects muscle tone. It is characterized by involuntary movements of the abdominal wall and rhythmic, repetitive contractions of the accessory or respiratory muscles, all of which are innervated by the cervical nerve roots. CASE DESCRIPTION: We reviewed the case of a 57-year-old male patient who underwent surgery for a left cerebellar hemorrhage. He exhibited persistent myoclonus in the palate, jaw, and thoracoabdominal region. Following treatment, there was a significant reduction in flutter amplitude in these areas. CONCLUSION: The clinical rarity and variability of presentations often make diagnosis challenging and delayed. It is believed that this condition stems from abnormal excitation within the central nervous system or neural pathways that involve the phrenic nerve. Another potential mechanism is the direct irritation of the diaphragm. Ultrasound, chest fluoroscopy, and electromyography (EMG) can support the diagnosis. Various pharmacological and surgical treatments have been tried, yet specific treatment guidelines are still lacking.

The reliability and validity of rapid transcranial magnetic stimulation mapping for muscles under active contraction.

Rapid mapping is a transcranial magnetic stimulation (TMS) mapping method which can significantly reduce data collection time compared to traditional approaches. However, its validity and reliability has only been established for upper-limb muscles during resting-state activity. Here, we determined the validity and reliability of rapid mapping for non-upper limb muscles that require active contraction during TMS: the masseter and quadriceps muscles. Eleven healthy participants attended two sessions, spaced two hours apart, each involving rapid and 'traditional' mapping of the masseter muscle and three quadriceps muscles (rectus femoris, vastus medialis, vastus lateralis). Map parameters included map volume, map area and centre of gravity (CoG) in the medial-lateral and anterior-posterior directions. Low to moderate measurement errors (%SEMeas = 10-32) were observed across muscles. Relative reliability varied from good-to-excellent (ICC = 0.63-0.99) for map volume, poor-to-excellent (ICC = 0.11-0.86) for map area, and fair-to-excellent for CoG (ICC = 0.25-0.8) across muscles. There was Bayesian evidence of equivalence (BF's > 3) in most map outcomes between rapid and traditional maps across all muscles, supporting the validity of the rapid mapping method. Overall, rapid TMS mapping produced similar estimates of map parameters to the traditional method, however the reliability results were mixed. As mapping of non-upper limb muscles is relatively challenging, rapid mapping is a promising substitute for traditional mapping, however further work is required to refine this method.

Development of an individualized stable and force-reducing lower-limb exoskeleton.

In this study, an individualized and stable passive-control lower-limb exoskeleton robot was developed. Users' joint angles and the center of pressure (CoP) of one of their soles were input into a convolutional neural network (CNN)-long short-term memory (LSTM) model to evaluate and adjust the exoskeleton control scheme. The CNN-LSTM model predicted the fitness of the control scheme and output the results to the exoskeleton robot, which modified its control parameters accordingly to enhance walking stability. The sole's CoP had similar trends during normal walking and passive walking with the developed exoskeleton; they-coordinates of the CoPs with and without the exoskeleton had a correlation of 91%. Moreover, electromyography signals from the rectus femoris muscle revealed that it exerted 40% less force when walking with a stable stride length in the developed system than when walking with an unstable stride length. Therefore, the developed lower-limb exoskeleton can be used to assist users in achieving balanced and stable walking with reduced force application. In the future, this exoskeleton can be used by patients with stroke and lower-limb weakness to achieve stable walking.