Search Results (78)

Robot learning from human demonstration pioneers an effective mapping paradigm for endowing robots with human-like operational capabilities. This paper proposes a bio-signal-guided robot adaptive stiffness learning framework grounded in the conclusion that muscle activation of the human arm is positively correlated with the endpoint stiffness. First, we propose a human-teleoperated demonstration platform enabling real-time modulation of robot end-effector stiffness by human tutors during operational tasks. Second, we develop a dual-stage probabilistic modeling architecture employing the Gaussian mixture model and Gaussian mixture regression to model the temporal–motion correlation and the motion–sEMG relationship, successively. Third, a real-world experiment was conducted to validate the effectiveness of the proposed skill transfer framework, demonstrating that the robot achieves online adaptation of Cartesian impedance characteristics in contact-rich tasks. This paper provides a simple and intuitive way to plan the Cartesian impedance parameters, transcending the classical method that requires complex human arm endpoint stiffness identification before human demonstration or compensation for the difference in human–robot operational effects after human demonstration. Full article

Muscle activity is composed of fast and slow activations. The detection of the onset time of the electromyogram signal, which is slowly activated, is difficult. This paper proposes a detection method based on marginal spectral entropy (MSE). The surface electromyography (sEMG) signal of the soleus during normal walking was collected by a wireless electromyography acquisition system. The proposed MSE-based detection method is based on the Hilbert–Huang transform (HHT) combined with information entropy. By comparing the changes in MSE before and after muscle activation to plot a trend line, the point of fastest change on the trend line was defined as the onset time of muscle activation. This method was compared with the amplitude threshold method and the Teager–Kaiser energy (TKE) operator method. The results show that the onset time of muscle activation detected by this method is 0.14 s earlier than the amplitude threshold method and 0.16 s earlier than the TKE operator method. The detection results were significantly different (p < 0.05), indicating that this method has higher detection accuracy for the onset time of the sEMG signal, which is slowly activated. Full article

This study aims to apply standard statistics and InterCriteria analysis (ICrA) for assessing the effects of different transcranial magnetic stimulation (TMS) intensities and three muscle conditions on the evoked responses of the first dorsal interosseous muscle (FDIM). Surface electromyograms from the right FDIM of ten right-handed healthy volunteers were recorded, and amplitudes of motor evoked potentials (MEPs), latencies of MEPs, and silent periods were obtained. ICrA was used for the first time as a supplementary tool along with the applied statistical methods. Three case studies were processed by the ICrA approach for a wide examination of neuromuscular excitability in humans. As a result, the relations between increasing TMS intensities, MEP amplitudes, MEP latencies, and silent periods were established at relaxed muscle condition, isometric index finger abduction condition, and co-contraction of antagonist muscles condition. Also, the dependencies between MEP amplitudes, MEP latencies, and silent periods themselves, and for different TMS intensities, were outlined. The results confirmed relations known from the literature and showed new ones. Full article

Deep neural networks (DNNs) and transfer learning (TL) have been used to improve surface electromyogram (sEMG)-based force estimation. However, prior studies focused mostly on applying TL within one joint, which limits dataset size and diversity. Herein, we investigated cross-joint TL between two upper-limb joints with four DNN architectures using sliding windows. We used two feedforward and two recurrent DNN models with feature engineering and feature learning, respectively. We found that the dependencies between sEMG and force are short-term (<400 ms) and that sliding windows are sufficient to capture them, suggesting that more complicated recurrent structures may not be necessary. Also, using DNN architectures reduced the required sliding window length. A model pre-trained on elbow data was fine-tuned on hand–wrist data, improving force estimation accuracy and reducing the required training data amount. A convolutional neural network with a 391 ms sliding window fine-tuned using 20 s of training data had an error of 6.03 ± 0.49% maximum voluntary torque, which is statistically lower than both our multilayer perceptron model with TL and a linear regression model using 40 s of training data. The success of TL between two distinct joints could help enrich the data available for future deep learning-related studies. Full article

Smart clothes equipped with flexible sensing systems provide a comfortable means to track health status in real time. Although these sensors are flexible and small, the core signal-processing units still rely on a conventional printed circuit board (PCB), making current health-monitoring devices bulky and inconvenient to wear. In this study, a printed fabric-based hybrid circuit was designed and prepared—with a series of characteristics, such as surface/sectional morphology, electrical properties, and stability—to study its reliability. Furthermore, to verify the function of the fabric-based circuit, simulations and measurements of the circuit, as well as the collection and processing of a normal adult’s electrophysiological signals, were conducted. Under 10,000 stretching and bending cycles with a certain elongation and bending angle, the resistance remained 0.27 Ω/cm and 0.64 Ω/cm, respectively, demonstrating excellent conductivity and reliability. Additionally, the results of the simulation and experiment showed that the circuit can successfully amplify weak electrocardiogram (ECG) signals with a magnification of 1600 times with environmental filtering and 50 Hz of industrial frequency interference. This technology can monitor human electrophysiological signals, such as ECGs, electromyograms (EMGs), and joint motion, providing valuable practical guidance for the unobtrusive monitoring of smart clothes. Full article

Bioelectrical signal measurements play a crucial role in clinical diagnosis and continuous health monitoring. Conventional wet electrodes, however, present limitations as they are conductive gel for skin irritation and/or have inflexibility. Here, we developed a cost-effective and user-friendly stretchable dry electrode constructed with a flexible network of Ag/AgCl nanowires embedded in polydimethylsiloxane (PDMS). We compared the performance of the stretched Ag/AgCl nanowire electrode with commonly used commercial wet electrodes to measure electrocardiogram (ECG), electromyogram (EMG), and electroencephalogram (EEG) signals. All the signal-to-noise ratios (SNRs) of the as-fabricated or stretched (50% tensile strain) Ag/AgCl nanowire electrodes are higher than that measured by commercial wet electrodes as well as other dry electrodes. The evaluation of ECG signal quality through waveform segmentation, the signal quality index (SQI), and heart rate variability (HRV) reveal that both the as-fabricated and stretched Ag/AgCl nanowire electrode produce high-quality signals similar to those obtained from commercial wet electrodes. The stretchable electrode exhibits high sensitivity and dependability in measuring EMG and EEG data, successfully capturing EMG signals associated with muscle activity and clearly recording α-waves in EEG signals during eye closure. Our stretchable dry electrode shows enhanced comfort, high sensitivity, and convenience for curved surface biosignal monitoring in clinical contexts. Full article

This study delves into decoding hand gestures using surface electromyography (EMG) signals collected via a precision Myo-armband sensor, leveraging machine learning algorithms. The research entails rigorous data preprocessing to extract features and labels from raw EMG data. Following partitioning into training and testing sets, four traditional machine learning models are scrutinized for their efficacy in classifying finger movements across seven distinct gestures. The analysis includes meticulous parameter optimization and five-fold cross-validation to evaluate model performance. Among the models assessed, the Random Forest emerges as the top performer, consistently delivering superior precision, recall, and F1-score values across gesture classes, with ROC-AUC scores surpassing 99%. These findings underscore the Random Forest model as the optimal classifier for our EMG dataset, promising significant advancements in healthcare rehabilitation engineering and enhancing human–computer interaction technologies. Full article

(1) Background: Subclinical hypothyroidism (SHT) is a condition that has been a subject of controversy in the literature due to its association with psychological and psychiatric symptoms as well as autonomic imbalances. To gain a better understanding of the effects of SHT on patients, a research study has been undertaken to investigate the presence of psychological symptoms and autonomic imbalances in a group of individuals diagnosed with SHT. (2) Methods: In this case–control study, 50 patients diagnosed with SHT who accessed the Department of Endocrinology of the University of Pisa were consecutively recruited. Psychological symptoms were measured through the Crown–Crisp Experiential Index (CCEI), whereas autonomic imbalance was described using the Psychophysiological Stress Profile (PSP), with simultaneous recording of the following psychophysiological parameters: Surface Electromyogram (sEMG), Skin Conductance Level (SCL), heart rate (HR), and peripheral temperature (PT). The patients’ values were compared to those of 50 healthy control subjects. (3) Results: The comparison between groups highlighted significant differences in the CCEI and PSP. In particular, patients reported higher rates of psychological symptoms (anxiety, depression, somatic complaints, and hysteria behavior). Significantly higher levels of autonomic arousal were also recorded. More specifically, the sEMG, SCL, HR, and PT values were different between the two groups. (4) Conclusions: The study has confirmed the presence of autonomic hyperarousal in patients diagnosed with subclinical hypothyroidism. This is likely due to the body’s attempt to compensate for a general lack of energy by accelerating the autonomic activity. The findings also underline the significance of a comprehensive assessment approach that takes into account various dimensions such as psychological and psychophysical well-being. Such an approach helps in evaluating the impact of subclinical diseases on overall health and well-being. Full article

The aim of this study was to compare seat pressure asymmetries before and after 30 min cycling at constant intensity in association with pelvic anthropometric parameters and skeletal muscle fatigue. Twelve male road cyclists aged 18–30 years (mean training experience 9.9 ± 2.5 years) participated. Pelvic anthropometric data and body composition were measured with dual-energy X-ray absorptiometry. Participants performed 30 min cycling at 50% peak power output at constant intensity on a cyclus-2 ergometer. Muscle fatigue during cycling was assessed by surface electromyogram spectral mean power frequency (MPF) for the back, gluteal, and thigh muscles. The pressure mapping system was used to assess sitting symmetry before and after the cycling exercise. At the end of cycling, MPF was decreased (p < 0.05) in the dominant side’s erector spinae muscle and the contralateral gluteal muscle. After the exercise, a significant (p < 0.05) asymmetry in seat pressure was observed under the ischial tuberosity based on the peak pressure right to left ratio, whereas peak pressure decreased under the left ischial tuberosity. After the exercise, the relationship (p < 0.05) between pelvis width and pressure under the ischial tuberosity occurred on the dominant side of the body. In conclusion, an asymmetry was revealed after the constant-load cycling exercise by peak pressure ratio right to left side. Further studies should address the role of seat pressure asymmetries before and after cycling exercises at different intensities and durations. Full article

Robot-assisted bilateral arm training has demonstrated its effectiveness in improving motor function in individuals post-stroke, showing significant enhancements with increased repetitions. However, prolonged training sessions may lead to both mental and muscle fatigue. We conducted two types of robot-assisted bimanual wrist exercises on 16 healthy adults, separated by one week: long-duration, low-resistance workouts and short-duration, high-resistance exercises. Various measures, including surface electromyograms, near-infrared spectroscopy, heart rate, and the Borg Rating of Perceived Exertion scale, were employed to assess fatigue levels and the impacts of exercise intensity. High-resistance exercise resulted in a more pronounced decline in electromyogram median frequency and recruited a greater amount of hemoglobin, indicating increased muscle fatigue and a higher metabolic demand to cope with the intensified workload. Additionally, high-resistance exercise led to increased sympathetic activation and a greater sense of exertion. Conversely, engaging in low-resistance exercises proved beneficial for reducing post-exercise muscle stiffness and enhancing muscle elasticity. Choosing a low-resistance setting for robot-assisted wrist movements offers advantages by alleviating mental and physiological loads. The reduced training intensity can be further optimized by enabling extended exercise periods while maintaining an approximate dosage compared to high-resistance exercises. Full article

Surface electromyogram (sEMG)-based gesture recognition has emerged as a promising avenue for developing intelligent prostheses for upper limb amputees. However, the temporal variations in sEMG have rendered recognition models less efficient than anticipated. By using cross-session calibration and increasing the amount of training data, it is possible to reduce these variations. The impact of varying the amount of calibration and training data on gesture recognition performance for amputees is still unknown. To assess these effects, we present four datasets for the evaluation of calibration data and examine the impact of the amount of training data on benchmark performance. Two amputees who had undergone amputations years prior were recruited, and seven sessions of data were collected for analysis from each of them. Ninapro DB6, a publicly available database containing data from ten healthy subjects across ten sessions, was also included in this study. The experimental results show that the calibration data improved the average accuracy by 3.03%, 6.16%, and 9.73% for the two subjects and Ninapro DB6, respectively, compared to the baseline results. Moreover, it was discovered that increasing the number of training sessions was more effective in improving accuracy than increasing the number of trials. Three potential strategies are proposed in light of these findings to enhance cross-session models further. We consider these findings to be of the utmost importance for the commercialization of intelligent prostheses, as they demonstrate the criticality of gathering calibration and cross-session training data, while also offering effective strategies to maximize the utilization of the entire dataset. Full article

Gait analysis has been studied over the last few decades as the best way to objectively assess the technical outcome of a procedure designed to improve gait. The treating physician can understand the type of gait problem, gain insight into the etiology, and find the best treatment with gait analysis. The gait parameters are the kinematics, including the temporal and spatial parameters, and lack the activity information of skeletal muscles. Thus, the gait analysis measures not only the three-dimensional temporal and spatial graphs of kinematics but also the surface electromyograms (sEMGs) of the lower limbs. Now, the shoe-worn GaitUp Physilog^® wearable inertial sensors can easily measure the gait parameters when subjects are walking on the general ground. However, it cannot measure muscle activity. The aim of this study is to measure the gait parameters using the sEMGs of the lower limbs. A self-made wireless device was used to measure the sEMGs from the vastus lateralis and gastrocnemius muscles of the left and right feet. Twenty young female subjects with a skeletal muscle index (SMI) below 5.7 kg/m² were recruited for this study and examined by the InBody 270 instrument. Four parameters of sEMG were used to estimate 23 gait parameters. They were measured using the GaitUp Physilog^® wearable inertial sensors with three machine learning models, including random forest (RF), decision tree (DT), and XGBoost. The results show that 14 gait parameters could be well-estimated, and their correlation coefficients are above 0.800. This study signifies a step towards a more comprehensive analysis of gait with only sEMGs. Full article

The analysis and recognition of sign languages are currently active fields of research focused on sign recognition. Various approaches differ in terms of analysis methods and the devices used for sign acquisition. Traditional methods rely on video analysis or spatial positioning data calculated using motion capture tools. In contrast to these conventional recognition and classification approaches, electromyogram (EMG) signals, which measure muscle electrical activity, offer potential technology for detecting gestures. These EMG-based approaches have recently gained attention due to their advantages. This prompted us to conduct a comprehensive study on the methods, approaches, and projects utilizing EMG sensors for sign language handshape recognition. In this paper, we provided an overview of the sign language recognition field through a literature review, with the objective of offering an in-depth review of the most significant techniques. These techniques were categorized in this article based on their respective methodologies. The survey discussed the progress and challenges in sign language recognition systems based on surface electromyography (sEMG) signals. These systems have shown promise but face issues like sEMG data variability and sensor placement. Multiple sensors enhance reliability and accuracy. Machine learning, including deep learning, is used to address these challenges. Common classifiers in sEMG-based sign language recognition include SVM, ANN, CNN, KNN, HMM, and LSTM. While SVM and ANN are widely used, random forest and KNN have shown better performance in some cases. A multilayer perceptron neural network achieved perfect accuracy in one study. CNN, often paired with LSTM, ranks as the third most popular classifier and can achieve exceptional accuracy, reaching up to 99.6% when utilizing both EMG and IMU data. LSTM is highly regarded for handling sequential dependencies in EMG signals, making it a critical component of sign language recognition systems. In summary, the survey highlights the prevalence of SVM and ANN classifiers but also suggests the effectiveness of alternative classifiers like random forests and KNNs. LSTM emerges as the most suitable algorithm for capturing sequential dependencies and improving gesture recognition in EMG-based sign language recognition systems. Full article

Disposable wearable sensors that ultrathin and conformable to the skin are of significant interest as affordable and easy-to-use devices for short-term recording. This study presents a facile and low-cost method for transferring spray-coated silver nanowire (AgNW) composite films onto human skin using glossy paper (GP) and liquid bandages (LB). Due to the moderately hydrophobic and rough surface of the GP, the ultrathin AgNWs composite film (~200 nm) was easily transferred onto human skin. The AgNW composite films conformally attached to the skin when applied with a LB, resulting in the stable and continuous recording of wearable electrophysiological signals, including electromyogram (EMG), electrocardiogram (ECG), and electrooculogram (EOG). The volatile LB, deposited on the skin via spray coating, promoted rapid adhesion of the transferred AgNW composite films, ensuring stability to the AgNWs in external environments. The AgNWs composite supported with the LB film exhibited high water vapor breathability (~28 gm⁻²h⁻¹), which can avoid the accumulation of sweat at the skin–sensor interface. This approach facilitates the creation of rapid, low-cost, and disposable tattoo-like sensors that are practical for extended use. Full article

Leg prostheses specially adapted for cycling in patients with transtibial amputation can be advantageous for recreational practice; however, their required features are not fully understood. Therefore, we aimed to evaluate the efficiency of unilateral cycling with a transtibial prosthesis and the characteristics of different attachment positions (middle and tip of the foot) between the prosthetic foot and the pedal. The cycling practice was performed on an ergometer at 40 W and 60 W resistance levels while participants (n = 8) wore custom-made orthoses to simulate prosthesis conditions. Using surface electromyogram, motion tracking, and power meter pedals, biomechanical data were evaluated and compared with data obtained through regular cycling. The results showed that power delivery became more asymmetrical at lower workloads for both orthosis conditions, while hip flexion and muscle activity of the knee extensor muscles in the sound leg increased. While both pedal attachment positions showed altered hip and knee joint angles for the leg wearing the orthosis, the middle of the foot attachment presented more symmetric power delivery. In conclusion, the middle of the foot attachment position presented better symmetry between the intact and amputated limbs during cycling performed for rehabilitation or recreation. Full article