Algorithmically detectable directional changes in upper extremity motion indicate substantial myoelectric shoulder muscle fatigue during a repetitive manual task

A New Approach to Quantifying Muscular Fatigue Using Wearable EMG Sensors during Surgery: An Ergonomic Case Study

(1) Background: Surgeons are exposed to musculoskeletal loads that are comparable to those of industrial workers. These stresses are harmful for the joints and muscles and can lead to musculoskeletal disorders (MSD) and working incapacity for surgeons. In this paper, we propose a novel ergonomic and visualization approach to assess muscular fatigue during surgical procedures. (2) Methods: The activity of eight muscles from the shoulder girdle and the cervical/lumbar spines were evaluated using position and electromyographic wearable sensors while a surgeon performed an arthroscopic rotator-cuff surgery on a patient. The time and frequency-domain variables of the root-mean-square amplitude and mean power frequency, respectively, were calculated from an electromyographic signal. (3) Results: The entire surgical procedure lasted 73 min and was divided into 10 sub-phases associated with specific level of muscular activity and fatigue. Most of the muscles showed activity above 60%, while the middle trapezius muscles were almost constantly activated (>20%) throughout the surgical procedure. (4) Conclusion: Wearable sensors can be used during surgical procedure to assess fatigue. Periods of low-to-high activity and fatigue can be evaluated and visualized during surgery. Micro-breaks throughout surgical procedures are suggested to avoid fatigue and to prevent the risk of developing MSD.

The effects of operating height and the passage of time on the end-point performance of fine manipulative tasks that require high accuracy

Sustained shoulder abduction, which results from an inappropriate worktable height or tool shape and long task hours, leads to an accumulation of muscle fatigue and subsequent work-related injuries in workers. It can be alleviated by controlling the table height or ergonomic tool design, but workers who are doing some types of work that require a discomfortable posture, such as minimally invasive surgery, cannot avoid these situations. Loads to the shoulder joint or muscles result in several problems, such as muscle fatigue, deterioration of proprioception or changing movement strategies of the central nervous system, and these are critical to work that requires a high accuracy of the upper extremities. Therefore, in this paper, we designed and conducted an experiment with human participants to discuss how an inappropriate height of the work-table affects the task performance of workers who are performing a fine manipulative task that requires high accuracy of the end point. We developed an apparatus that can control the height and has four touch screens to evaluate the end-point accuracy with two different heights. Eighteen adults (9 women and 9 men) participated in the experiments, and the electromyography of their shoulder muscles, their movement stability, and task performance were measured for the analysis. We found that inappropriate height of a table brings about muscle fatigue, and time elapsed for conducting tasks accelerated the phenomenon. Task performance deteriorated according to increased fatigue, and improved movement stability is not enough to compensate for these situations.

Exploring Localized Muscle Fatigue Responses at Current Upper-Extremity Ergonomics Threshold Limit Values

OBJECTIVE

The purpose of this study was to evaluate localized muscle fatigue responses at three upper-extremity ergonomics threshold limit value (TLV) duty cycles.

BACKGROUND

Recently, a TLV equation was published to help mitigate excessive development of localized muscle fatigue in repetitive upper limb tasks. This equation predicts acceptable levels of maximal voluntary contraction (% MVC) for a given duty cycle (DC). Experimental validation of this TLV curve has not yet been reported, which can help guide utilization by practitioners.

METHOD

Eighteen participants performed intermittent isometric elbow flexion efforts, in three separate counter-balanced sessions, at workloads defined by the American Conference of Governmental Industrial Hygenists' (ACGIH) TLV equation: low DC (20% DC, 29.6% MVC), medium DC (40% DC, 19.7% MVC), and high DC (60% DC, 13.9% MVC). Targeted localized muscle fatigue (LMF) of the biceps brachii was tracked across numerous response variables, including decline in strength (MVC), electromyography (EMG) amplitude and mean power frequency (MnPF), and several psychophysical ratings.

RESULTS

At task completion, biceps MnPF and MVC (strength) were significantly different between each TLV workload, with the high DC condition eliciting the largest declines in MnPF and MVC.

CONCLUSION

Findings demonstrate that working at different DCs along the ACGIH TLV curve may not be equivalent in preventing excessive LMF. Higher DC workloads elicited a greater LMF response across several response variables.

APPLICATION

High DC work of the upper extremity should be avoided to mitigate excess LMF development. Current TLVs for repetitive upper-extremity work may overestimate acceptable relative contraction thresholds, particularly at higher duty cycles.

InverseMuscleNET: Alternative Machine Learning Solution to Static Optimization and Inverse Muscle Modeling

InverseMuscleNET, a machine learning model, is proposed as an alternative to static optimization for resolving the redundancy issue in inverse muscle models. A recurrent neural network (RNN) was optimally configured, trained, and tested to estimate the pattern of muscle activation signals. Five biomechanical variables (joint angle, joint velocity, joint acceleration, joint torque, and activation torque) were used as inputs to the RNN. A set of surface electromyography (EMG) signals, experimentally measured around the shoulder joint for flexion/extension, were used to train and validate the RNN model. The obtained machine learning model yields a normalized regression in the range of 88-91% between experimental data and estimated muscle activation. A sequential backward selection algorithm was used as a sensitivity analysis to discover the less dominant inputs. The order of most essential signals to least dominant ones was as follows: joint angle, activation torque, joint torque, joint velocity, and joint acceleration. The RNN model required 0.06 s of the previous biomechanical input signals and 0.01 s of the predicted feedback EMG signals, demonstrating the dynamic temporal relationships of the muscle activation profiles. The proposed approach permits a fast and direct estimation ability instead of iterative solutions for the inverse muscle model. It raises the possibility of integrating such a model in a real-time device for functional rehabilitation and sports evaluation devices with real-time estimation and tracking. This method provides clinicians with a means of estimating EMG activity without an invasive electrode setup.

Continuous, integrated sensors for predicting fatigue during non-repetitive work: demonstration of technique in the operating room

Surface electromyography (sEMG) can monitor muscle activity and potentially predict fatigue in the workplace. However, objectively measuring fatigue is challenging in complex work with unpredictable work cycles where sEMG may be influenced by the dynamically changing posture demands. This study proposes a multi-modal approach integrating sEMG with motion sensors and demonstrates the approach in the live surgical work environment. Seventy-two exposures from twelve participants were collected, including self-reported musculoskeletal discomfort, sEMG, and postures. Posture sensors were used to identify time windows where the surgeon was static and in non-demanding positions, and mean power frequencies (MPF) were then calculated during those time windows. In 57 out of 72 exposures (80%), participants experienced an increase in musculoskeletal discomfort. Integrated (multi-modality) measurements showed better performance than single-modality (sEMG) measurements in detecting decreases in MPF, a predictor of fatigue. Based on self-reported musculoskeletal discomfort, sensor-based thresholds for identifying fatigue are proposed for the trapezius and deltoid muscle groups. Practitioner summary Work-related fatigue is one of the intermediate risk factors to musculoskeletal disorders. This article presents an objective integrated approach to identify musculoskeletal fatigue using wearable sensors. The presented approach could be implemented by ergonomists to identify musculoskeletal fatigue more accurately and in a variety of workplaces. Abbreviations: sEMG: surface electromyography; IMU: inertia measurement unit; MPF: mean power frequency; ACGIH: American Conference of Governmental Industrial Hygienists; SAGES: Society of American Gastrointestinal and Endoscopic Surgeons; LD: left deltoid; LT: left trapezius; RD: right deltoid; RT: right trapezius.

MuscleNET: mapping electromyography to kinematic and dynamic biomechanical variables by machine learning

Objective.This paper proposes machine learning models for mapping surface electromyography (sEMG) signals to regression of joint angle, joint velocity, joint acceleration, joint torque, and activation torque.Approach.The regression models, collectively known as MuscleNET, take one of four forms: ANN (forward artificial neural network), RNN (recurrent neural network), CNN (convolutional neural network), and RCNN (recurrent convolutional neural network). Inspired by conventional biomechanical muscle models, delayed kinematic signals were used along with sEMG signals as the machine learning model's input; specifically, the CNN and RCNN were modeled with novel configurations for these input conditions. The models' inputs contain either raw or filtered sEMG signals, which allowed evaluation of the filtering capabilities of the models. The models were trained using human experimental data and evaluated with different individual data.Main results.Results were compared in terms of regression error (using the root-mean-square) and model computation delay. The results indicate that the RNN (with filtered sEMG signals) and RCNN (with raw sEMG signals) models, both with delayed kinematic data, can extract underlying motor control information (such as joint activation torque or joint angle) from sEMG signals in pick-and-place tasks. The CNNs and RCNNs were able to filter raw sEMG signals.Significance.All forms of MuscleNET were found to map sEMG signals within 2 ms, fast enough for real-time applications such as the control of exoskeletons or active prostheses. The RNN model with filtered sEMG and delayed kinematic signals is particularly appropriate for applications in musculoskeletal simulation and biomechatronic device control.

Joint moment trade-offs across the upper extremity and trunk during repetitive work

Individuals can coordinate small kinematic changes at several degrees of freedom simultaneously in the presence of fatigue, leaving it unclear how overall biomechanical demands at each joint are altered. The purpose of this study was to evaluate trade-offs in joint moments between the trunk, shoulder, and elbow during repetitive upper extremity work. Participants performed four simulated workplace tasks cyclically until meeting fatigue termination criteria. Emergent fatigue-induced adaptations to repetitive work resulted in task-dependent trade-offs in joint moments. In general, reduced shoulder moments were compensated for by increased elbow and trunk joint moment contributions. Although mean joint moment changes were modest (range: 1-3 Nm) across participants, a wide distribution of responses was observed, with standard deviations exceeding 10 Nm. Re-distributing biomechanical demands across joints may alleviate constant tissue loads and facilitate continued task performance with fatigue but may be at the expense of increasing demands at adjacent joints.

Muscle Fatigue in the Three Heads of Triceps Brachii During Intensity and Speed Variations of Triceps Push-Down Exercise

The objective of this study was to investigate the effects of changes in exercise intensity and speed on the three heads of the triceps brachii (TB) during triceps push-down exercise until task failure. Twenty-five subjects performed triceps push-down exercise at three different intensities (30, 45, and 60% 1RM) and speeds (slow, medium, and fast) until failure, and surface electromyography (sEMG) signals were recorded from the lateral, long and medial heads of the TB. The endurance time (ET), number of repetitions (NR) and rate of fatigue (ROF) were analyzed. Subsequently, the root-mean-square (RMS), mean power frequency (MPF) and median frequency (MDF) under no-fatigue (NF) and fatigue (Fa) conditions were statistically compared. The findings reveal that ROF increases with increase in the intensity and speed, and the opposite were obtained for the ET. The ROF in the three heads were comparable for all intensities and speeds. The ROF showed a significant difference (P < 0.05) among the three intensities and speeds for all heads. The three heads showed significantly different (P < 0.05) MPF and MDF values for all the performed exercises under both conditions, whereas the RMS values were significantly different only under Fa conditions. The current observations suggest that exercise intensity and speed affect the ROF while changes in intensity do not affect the MPF and MDF under Fa conditions. The behavior of the spectral parameters indicate that the three heads do not work in unison under any of the conditions. Changes in the speed of triceps push-down exercise affects the lateral and long heads, but changes in the exercise intensity affected the attributes of all heads to a greater extent.