Application of singular spectrum-based change-point analysis to EMG-onset detection

A systematic investigation of detectors for low signal-to-noise ratio EMG signals

Background

Active participation of stroke survivors during robot-assisted movement therapy is essential for sensorimotor recovery. Robot-assisted therapy contingent on movement intention is an effective way to encourage patients' active engagement. For severely impaired stroke patients with no residual movements, a surface electromyogram (EMG) has been shown to be a viable option for detecting movement intention. Although numerous algorithms for EMG detection exist, the detector with the highest accuracy and lowest latency for low signal-to-noise ratio (SNR) remains unknown.

Methods

This study, therefore, investigates the performance of 13 existing EMG detection algorithms on simulated low SNR (0dB and -3dB) EMG signals generated using three different EMG signal models: Gaussian, Laplacian, and biophysical model. The detector performance was quantified using the false positive rate (FPR), false negative rate (FNR), and detection latency. Any detector that consistently showed FPR and FNR of no more than 20%, and latency of no more than 50ms, was considered an appropriate detector for use in robot-assisted therapy.

Results

The results indicate that the Modified Hodges detector - a simplified version of the threshold-based Hodges detector introduced in the current study - was the most consistent detector across the different signal models and SNRs. It consistently performed for ~90% and ~40% of the tested trials for 0dB and -3dB SNR, respectively. The two statistical detectors (Gaussian and Laplacian Approximate Generalized Likelihood Ratio) and the Fuzzy Entropy detectors have a slightly lower performance than Modified Hodges.

Conclusions

Overall, the Modified Hodges, Gaussian and Laplacian Approximate Generalized Likelihood Ratio, and the Fuzzy Entropy detectors were identified as the potential candidates that warrant further investigation with real surface EMG data since they had consistent detection performance on low SNR EMG data.

Exploring the EMG transient: the muscular activation sequences used as novel time-domain features for hand gestures classification

Introduction

Muscular activation sequences have been shown to be suitable time-domain features for classification of motion gestures. However, their clinical application in myoelectric prosthesis control was never investigated so far. The aim of the paper is to evaluate the robustness of these features extracted from the EMG signal in transient state, on the forearm, for classifying common hand tasks.

Methods

The signal associated to four hand gestures and the rest condition were acquired from ten healthy people and two persons with trans-radial amputation. A feature extraction algorithm allowed for encoding the EMG signals into muscular activation sequences, which were used to train four commonly used classifiers, namely Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), Non-linear Logistic Regression (NLR) and Artificial Neural Network (ANN). The offline performances were assessed with the entire sample of recruited people. The online performances were assessed with the amputee subjects. Moreover, a comparison of the proposed method with approaches based on the signal envelope in the transient state and in the steady state was conducted.

Results

The highest performance were obtained with the NLR classifier. Using the sequences, the offline classification accuracy was higher than 93% for healthy and amputee subjects and always higher than the approach with the signal envelope in transient state. As regards the comparison with the steady state, the performances obtained with the proposed method are slightly lower (<4%), but the classification occurred at least 200 ms earlier. In the online application, the motion completion rate reached up to 85% of the total classification attempts, with a motion selection time that never exceeded 218 ms.

Discussion

Muscular activation sequences are suitable alternatives to the time-domain features commonly used in classification problems belonging to the sole EMG transient state and could be potentially exploited in control strategies of myoelectric prosthesis hands.

Review of electromyography onset detection methods for real-time control of robotic exoskeletons

BACKGROUND

Electromyography (EMG) is a classical technique used to record electrical activity associated with muscle contraction and is widely applied in Biomechanics, Biomedical Engineering, Neuroscience and Rehabilitation Robotics. Determining muscle activation onset timing, which can be used to infer movement intention and trigger prostheses and robotic exoskeletons, is still a big challenge. The main goal of this paper was to perform a review of the state-of-the-art of EMG onset detection methods. Moreover, we compared the performance of the most commonly used methods on experimental EMG data.

METHODS

A total of 156 papers published until March 2022 were included in the review. The papers were analyzed in terms of application domain, pre-processing method and EMG onset detection method. The three most commonly used methods [Single (ST), Double (DT) and Adaptive Threshold (AT)] were applied offline on experimental intramuscular and surface EMG signals obtained during contractions of ankle and knee joint muscles.

RESULTS

Threshold-based methods are still the most commonly used to detect EMG onset. Compared to ST and AT, DT required more processing time and, therefore, increased onset timing detection, when applied on experimental data. The accuracy of these three methods was high (maximum error detection rate of 7.3%), demonstrating their ability to automatically detect the onset of muscle activity. Recently, other studies have tested different methods (especially Machine Learning based) to determine muscle activation onset offline, reporting promising results.

CONCLUSIONS

This study organized and classified the existing EMG onset detection methods to create consensus towards a possible standardized method for EMG onset detection, which would also allow more reproducibility across studies. The three most commonly used methods (ST, DT and AT) proved to be accurate, while ST and AT were faster in terms of EMG onset detection time, especially when applied on intramuscular EMG data. These are important features towards movement intention identification, especially in real-time applications. Machine Learning methods have received increased attention as an alternative to detect muscle activation onset. However, although several methods have shown their capability offline, more research is required to address their full potential towards real-time applications, namely to infer movement intention.

Ankle reaction times with tray usage following a slip perturbation between subjects with and without chronic low back pain

BACKGROUND

Abnormal stepping strategies have been associated with handheld tasks in subjects with chronic low back pain (LBP). However, the dominant ankle reactions of subjects with LBP remain unclear following a perturbation during handheld tasks.

RESEARCH QUESTION

Are there differences in the reaction times of the ankle muscles during handheld tasks between subjects with and without LBP following a treadmill-induced slip perturbation?

METHODS

Thirty-seven right limb dominant subjects with LBP and 37 subjects without LBP participated in the study. Each subject was introduced to a slip perturbation (1.37 m/sec velocity for 8.22 cm) with and without a handheld tray in random order. Subjects were allowed to recover by stepping forward for a 0.12 s duration while bilateral tibialis anterior (TA) and gastrocnemius (GA) muscle reaction times were measured by electromyography (EMG).

RESULTS

The EMG results indicated that the groups demonstrated significant interactions on the limb sides and muscles (F = 4.86, p = 0.03). The dominant TA reaction time was significantly faster in the LBP group (t = 2.14, p = 0.03) while holding a tray.

SIGNIFICANCE

The LBP group demonstrated faster reaction times on the dominant TA muscles during perturbations. Clinicians need to consider dominance-dependent compensatory ankle dorsiflexion strategies in LBP patients to help enhance dynamic balance and control.

The reaction times and symmetry indices in the bilateral trunk and limb muscles in control subjects and subjects with low back pain that persisted two months or longer

PURPOSE

This study was conducted to investigate the reaction times and symmetry index (SI) of the bilateral trunk and limb muscles between control subjects and subjects with low back pain (LBP) that persisted for two months or longer.

METHODS

Fifty-seven right limb dominant subjects (31 healthy control subjects and 26 subjects with LBP) participated in this study. The subjects were exposed to a slip perturbation (0.24 m/sec velocity for 1.20 cm), which caused them to move forward for 0.10 s in standing while holding a tray. The electromyography (EMG) electrodes were placed on the bilateral erector spinae (ES), rectus abdominis (RA), rectus femoris, hamstring, tibialis anterior, gastrocnemius, biceps brachii (BB), and triceps brachii muscles. The reaction times were analyzed, and the SI was used to compare the bilateral trunk and limb muscles for the degree of asymmetry between groups.

RESULTS

The ES reaction time was significantly delayed in the control group (0.33 ± 0.22 vs. 0.22 ± 0.17; t = 2.25, p = 0.03). The SI of reaction times was significantly different on the RA (t = -2.28, p = 0.03), ES (t = -2.36, p = 0.04), and BB (t = -2.15, p = 0.04) muscles between groups.

CONCLUSION

The delayed non-dominant ES reaction time might indicate a freedom of pain recurrence in the control group. Although the asymmetry increased on the RA and BB muscles in the LBP group, it decreased on the ES muscle. The asymmetries on the trunk and BB muscles were evident in the LBP group. The asymmetrical reactions in the arm-trunk muscles need to be considered for rehabilitation strategies.

[Changes in corneal nerves fibers in the early stages of Parkinson's disease according to in vivo confocal microscopy (preliminary report)]

One of the research directions of the so-called non-motor manifestations of Parkinson's disease (PD) is associated with the assessment of structural and functional changes in the organ of vision. An assessment of the state of thin non-myelinated corneal nerve fibers (CNF) in Parkinson's disease seems to be promising considering the neurodegenerative nature of the disease, as well as the possibility of objective intravital assessment of both functional and structural changes in CNF.

PURPOSE

To analyze the changes in the course and structure of corneal nerve fibers in the early stages of Parkinson's disease based on an objective algorithm of in vivo corneal confocal microscopy (CCM).

MATERIAL AND METHODS

The study was conducted on a group of 16 patients aged 39 to 66 years with verified diagnosis of PD. In addition to standard neurological and ophthalmological examinations, all patients underwent IVCCM on a Heidelberg Retinal Tomograph device with special Rostock Cornea Module (HRT3 RCM), followed by processing of the obtained images using a uniquely designed analysis algorithm.

RESULTS

A significant decrease in the directional anisotropy coefficient and an increase in the directional symmetry coefficient of the nerve fibers of the cornea were established (average values 3.15±1.08 and 0.92±0.04, respectively); in healthy individuals of the identical age range these indicators are 3.5±0.85 and 0.86±0.11, respectively. In addition, qualitative structural changes were noted, which consisted of an increase in the number of branches from the main nerve trunks, an increase in the tortuosity of CNF, multidirectionality, and "beaded" shape. In 9 cases, the presence of macrophages was revealed - dendritic Langerhans cells, which is an indirect sign of the inflammatory process.

CONCLUSION

The preliminary nature of the results obtained in this study and the need for further research in this area are related, on the one hand, to a small sample of observations and, on the other hand, to the criterion used to assess the status of CNF based on a comparative analysis with conditionally normal indicators. In the future, in order to solve the problem of the uniqueness of changes in CNF and the possibility of using these changes as a marker for PD progression, longitudinal studies are required to reveal the presence or absence of a correlation between the stage of the disease, the results of known monitoring methods (e.g. electromyography) and quantitative indicators of the status of CNF.

Internal consistencies of the delayed trunk muscle reaction times following a treadmill-induced slip perturbation while holding and not holding a tray

BACKGROUND

Reaction time task performance using electromyography (EMG) has been widely studied in the evaluation of motor responses. However, specific testing conditions with tray usage and the reliability of the bilateral trunk muscle reactions have not been proven.

RESEARCH QUESTIONS

Are there internal consistencies of the reaction times for a particular condition, such as a handheld task, among the examiners? Is there a delayed reaction time on the dominant abdominal muscle in response to a treadmill-induced slip perturbation while holding or not holding a tray?

METHODS

One hundred and nineteen right upper and lower limb dominant individuals (71 female and 48 male subjects) were exposed to a treadmill-induced slip perturbation (0.24 m/s velocity for 1.2 cm) for 0.10 s in standing. The EMG electrodes were placed on both sides of the rectus abdominis (RA) and erector spinae (ES) muscles. The reliability of the test was established by using Cronbach's alpha, intra-class correlation coefficients (ICC_{2, k}), and the standard error of measurements.

RESULTS

The results for holding a tray indicated a high degree of consistency based on Cronbach's alpha for the left RA (0.79), right RA (0.86), left ES (0.82), and right ES (0.73) muscles. However, there was a significant reaction time difference among trunk muscles (F = 10.58, p = 0.002) while not holding a tray. The post-hoc results indicated that the right RA muscle was delayed more than the bilateral ES muscles, although there was no significant difference with the left RA muscle.

SIGNIFICANCE

Overall, the EMG analyses for the reaction times were highly consistent with and without tray usage. The reaction times of the dominant abdominal muscles were delayed while not holding a tray. Given the high reliability, compensatory strategies by trunk dominance might be considered with a tray usage task.

Investigating the performance of an amplitude-independent algorithm for detecting the hand muscle activity of stroke survivors

To make robotic hand devices controlled by surface electromyography (sEMG) signals feasible and practical tools for assisting patients with hand impairments, the problems that prevent these devices from being widely used have to be overcome. The most significant problem is the involuntary amplitude variation of the sEMG signals due to the movement of electrodes during forearm motion. Moreover, for patients who have had a stroke or another neurological disease, the muscle activity of the impaired hand is weak and has a low signal-to-noise ratio (SNR). Thus, muscle activity detection methods intended for controlling robotic hand devices should not depend mainly on the amplitude characteristics of the sEMG signal in the detection process, and they need to be more reliable for sEMG signals that have a low SNR. Since amplitude-independent muscle activity detection methods meet these requirements, this paper investigates the performance of such a method on people who have had a stroke in terms of the detection of weak muscle activity and resistance to false alarms caused by the involuntary amplitude variation of sEMG signals; these two parameters are very important for achieving the reliable control of robotic hand devices intended for people with disabilities. A comparison between the performance of an amplitude-independent muscle activity detection algorithm and three amplitude-dependent algorithms was conducted by using sEMG signals recorded from six hemiparesis stroke survivors and from six healthy subjects. The results showed that the amplitude-independent algorithm performed better in terms of detecting weak muscle activity and resisting false alarms.

A new detection method for EMG activity monitoring

This paper introduces a new approach for electromyography (EMG) activity monitoring based on an improved version of the adaptive linear energy detector (ALED), a widely used technique in voice activity detection. More precisely, we propose a modified ALED technique (named M-ALED) to improve the method's robustness with respect to noise. To achieve this objective, M-ALED relies on the Teager-Kaiser operator for signal pre-conditioning to increase the SNR and uses the order statistics to gain robustness against the signal's impulsiveness. We propose again to exploit the order statistics for the initial signal baseline estimation to deal with the cases where such information is unavailable. Finally, since M-ALED detects the signal's activity at the frame level, we propose in a second stage to refine this detection (at the sample level) by using a constant false alarm rate (CFAR) approach leading to the fine M-ALED (FM-ALED) solution. The performance of FM-ALED is assessed via real and synthetic EMG signal recordings and the obtained results highlight its effectiveness as compared with the state-of-the-art methods (it reduces the mean error probability by a factor close to 2).

Wearable Monitoring Devices for Biomechanical Risk Assessment at Work: Current Status and Future Challenges-A Systematic Review

Background: In order to reduce the risk of work-related musculoskeletal disorders (WMSDs) several methods have been developed, accepted by the international literature and used in the workplace. The purpose of this systematic review was to describe recent implementations of wearable sensors for quantitative instrumental-based biomechanical risk assessments in prevention of WMSDs. Methods: Articles written until 7 May 2018 were selected from PubMed, Scopus, Google Scholar and Web of Science using specific keywords. Results: Instrumental approaches based on inertial measurement units and sEMG sensors have been used for direct evaluations to classify lifting tasks into low and high risk categories. Wearable sensors have also been used for direct instrumental evaluations in handling of low loads at high frequency activities by using the local myoelectric manifestation of muscle fatigue estimation. In the field of the rating of standard methods, on-body wireless sensors network-based approaches for real-time ergonomic assessment in industrial manufacturing have been proposed. Conclusions: Few studies foresee the use of wearable technologies for biomechanical risk assessment although the requirement to obtain increasingly quantitative evaluations, the recent miniaturization process and the need to follow a constantly evolving manual handling scenario is prompting their use.

A novel approach to automatically quantify the level of coincident activity between EMG and MMG signals

Although previous studies have highlighted both similarities and differences between the timing of electromyography (EMG) and mechanomyography (MMG) activities of muscles, there is no method to systematically quantify the temporal alignment between corresponding EMG and MMG signals. We proposed a novel method to determine the level of coincident activity in quasi-periodic MMG and EMG signals. The method optimizes 3 muscle-specific parameters: amplitude threshold, window size and minimum percent of EMG and MMG overlap using a particle swarm optimization algorithm to maximize the agreement (balanced accuracy) between electrical and mechanical muscle activity. The method was applied to bilaterally recorded EMG and MMG signals from 4 lower limb muscles per side of 25 pediatric participants during self-paced gait. Mean balanced accuracy exceeded 75% for all muscles except the lateral gastrocnemius, where EMG and MMG misalignment was notable (56% balanced accuracy). The proposed method can be applied to the criterion-driven comparison of simultaneously recorded myographic signals from two different measurement modalities during a motor task.

Is EMG a Viable Alternative to BCI for Detecting Movement Intention in Severe Stroke?

OBJECTIVE

In light of the shortcomings of current restorative brain-computer interfaces (BCI), this study investigated the possibility of using EMG to detect hand/wrist extension movement intention to trigger robot-assisted training in individuals without residual movements.

METHODS

We compared movement intention detection using an EMG detector with a sensorimotor rhythm based EEG-BCI using only ipsilesional activity. This was carried out on data of 30 severely affected chronic stroke patients from a randomized control trial using an EEG-BCI for robot-assisted training.

RESULTS

The results indicate the feasibility of using EMG to detect movement intention in this severely handicapped population; probability of detecting EMG when patients attempted to move was higher (p 0.001) than at rest. Interestingly, 22 out of 30 (or 73%) patients had sufficiently strong EMG in their finger/wrist extensors. Furthermore, in patients with detectable EMG, there was poor agreement between the EEG and EMG intent detectors, which indicates that these modalities may detect different processes.

CONCLUSION

A substantial segment of severely affected stroke patients may benefit from EMG-based assisted therapy. When compared to EEG, a surface EMG interface requires less preparation time, which is easier to don/doff, and is more compact in size.

SIGNIFICANCE

This study shows that a large proportion of severely affected stroke patients have residual EMG, which yields a direct and practical way to trigger robot-assisted training.

Automated event detection algorithm for two squatting protocols

INTRODUCTION

Squatting biomechanics assessed using motion analysis relies on the identification of specific events: start of descent, transition between descent/ascent and end of ascent. Automated identification reduces the time needed to process trials while allowing consistency across studies. The purpose of this study was to develop criteria for the identification of events and apply them to two squatting protocols in pathological patient and typically developing (TD) groups.

METHODS

Thirty-four subjects with hip dysplasia and 41 TD subjects were enrolled in this study. While instrumented with a full-body Plug-In-Gait marker set, participants performed two squatting protocols: a hold squat, where subjects paused for a count of three at their lowest squat depth, and a traditional squat, where the descent phase was immediately followed by the ascent phase. Reviewers analyzed the kinematic/kinetic waveforms of a subset of trials to develop criteria for events. Sagittal plane knee and vertical center of mass velocities were used to identify events and absolute vs. relative thresholds of the peak knee velocity were compared. These criteria were incorporated into an automatic event detection code.

RESULTS

Using a relative threshold algorithm, events were automatically identified in 244 of 259 total trials (94%). For the trials requiring manual placement of events (n=15 trials), there was perfect inter-rater reliability between research personnel.

CONCLUSIONS

The criteria developed for the automatic detection of squatting events was highly successful for both protocols in each participant group and was also highly reliable for research personnel to follow in the few instances where manual placement was necessary.

Effects of prone trunk extension exercise using different fixations and with and without abdominal drawing-in maneuver in healthy individuals

[Purpose] The purpose of this study was to investigate the differential effects of fixation of the thoracolumbar junction on the activity of the thoracic and lumbar muscle groups with and without the abdominal drawing-in maneuver (ADIM). [Subjects and Methods] The participants were 40 healthy adults. During trunk extension, thoracic and lumbar muscle activities were assessed using surface electromyography when fixing the pelvic and popliteal areas and the thoracolumbar junction, with and without the ADIM. [Results] The activity of the thoracic extensors at the T9 level applying thoracolumbar fixation with ADIM was significantly higher than with only pelvic fixation or pelvic fixation with ADIM during prone trunk extension. However, the activity of the lumbar extensors at the L3 level with pelvic fixation alone, without ADIM, was significantly higher than pelvic fixation with ADIM or additional thoracolumbar fixation with ADIM during prone trunk extension [Conclusion] This study demonstrated that fixation of the thoracolumbar junction with ADIM during trunk extension was more effective at eliciting thoracic extensor and minimising lumbar extensor activity than other conventional fixation methods.

A Fast Framework for Abrupt Change Detection Based on Binary Search Trees and Kolmogorov Statistic

Change-Point (CP) detection has attracted considerable attention in the fields of data mining and statistics; it is very meaningful to discuss how to quickly and efficiently detect abrupt change from large-scale bioelectric signals. Currently, most of the existing methods, like Kolmogorov-Smirnov (KS) statistic and so forth, are time-consuming, especially for large-scale datasets. In this paper, we propose a fast framework for abrupt change detection based on binary search trees (BSTs) and a modified KS statistic, named BSTKS (binary search trees and Kolmogorov statistic). In this method, first, two binary search trees, termed as BSTcA and BSTcD, are constructed by multilevel Haar Wavelet Transform (HWT); second, three search criteria are introduced in terms of the statistic and variance fluctuations in the diagnosed time series; last, an optimal search path is detected from the root to leaf nodes of two BSTs. The studies on both the synthetic time series samples and the real electroencephalograph (EEG) recordings indicate that the proposed BSTKS can detect abrupt change more quickly and efficiently than KS, t-statistic (t), and Singular-Spectrum Analyses (SSA) methods, with the shortest computation time, the highest hit rate, the smallest error, and the highest accuracy out of four methods. This study suggests that the proposed BSTKS is very helpful for useful information inspection on all kinds of bioelectric time series signals.

Influence of amplitude cancellation on the accuracy of determining the onset of muscle activity from the surface electromyogram

The purpose of the study was to quantify the influence of amplitude cancellation on the accuracy of detecting the onset of muscle activity based on an analysis of simulated surface electromyographic (EMG) signals. EMG activity of a generic lower limb muscle was simulated during the stance phase of human gait. Surface EMG signals were generated with and without amplitude cancellation by summing simulated motor unit potentials either before (cancellation EMG) or after (no-cancellation EMG) the potentials had been rectified. The two sets of EMG signals were compared at forces of 30% and 80% of maximum voluntary contraction (MVC) and with various low-pass filter cut-off frequencies. Onset time was determined both visually and by an algorithm that identified when the mean amplitude of the signal within a sliding window exceeded a specified standard deviation (SD) above the baseline mean. Onset error was greater for the no-cancellation conditions when determined automatically and by visual inspection. However, the differences in onset error between the two cancellation conditions appear to be clinically insignificant. Therefore, amplitude cancellation does not appear to limit the ability to detect the onset of muscle activity from the surface EMG.

Human behavior integration improves classification rates in real-time BCI

Brain-computer interfaces (BCI) offer potential for individuals with a variety of motor and sensory disabilities to interact with their environment, communicate and control mobility aids. Two key factors which affect the performance of a BCI and its usability are the feedback given to the participant and the subject's motivation. This paper presents the results from a study investigating the effects of feedback and motivation on the performance of the Strathclyde Brain Computer Interface. The paper discusses how the performance of the system can be improved by behavior integration and human-in-the-loop design.