Biomechanical Risk Classification in Repetitive Lifting Using Multi-Sensor Electromyography Data, Revised National Institute for Occupational Safety and Health Lifting Equation, and Deep Learning

Biosensors, Journal Year: 2025, Volume and Issue: 15(2), P. 84 - 84

Published: Feb. 1, 2025

Repetitive lifting tasks in occupational settings often result shoulder injuries, impacting both health and productivity. Accurately assessing the biomechanical risk of these remains a significant challenge ergonomics, particularly within manufacturing environments. Traditional assessment methods frequently rely on subjective reports limited observations, which can introduce bias yield incomplete evaluations. This study addresses limitations by generating utilizing comprehensive dataset containing detailed time-series electromyography (EMG) data from 25 participants. Using high-precision wearable sensors, EMG were collected eight muscles as participants performed repetitive tasks. For each task, index was calculated using revised National Institute for Occupational Safety Health (NIOSH) equation (RNLE). Participants completed cycles low-risk high-risk four-minute period, allowing muscle performance under realistic working conditions. extensive dataset, comprising over 7 million points sampled at approximately 1259 Hz, leveraged to develop deep learning models classify risk. To provide actionable insights practical ergonomics assessments, statistical features extracted raw data. Three models, Convolutional Neural Networks (CNNs), Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM), employed analyze predict level. The CNN model achieved highest performance, with precision 98.92% recall 98.57%, proving its effectiveness real-time assessments. These findings underscore importance aligning architectures characteristics optimize management. By integrating sensors this enables precise, real-time, dynamic significantly enhancing workplace safety protocols. approach has potential improve planning reduce incidence severity work-related musculoskeletal disorders, ultimately promoting better outcomes across various settings.

Language: Английский

---

The Effect of Sensor Feature Inputs on Joint Angle Prediction across Simple Movements

Sensors, Journal Year: 2024, Volume and Issue: 24(11), P. 3657 - 3657

Published: June 5, 2024

The use of wearable sensors, such as inertial measurement units (IMUs), and machine learning for human intent recognition in health-related areas has grown considerably. However, there is limited research exploring how IMU quantity placement affect movement prediction (HMIP) at the joint level. objective this study was to analyze various combinations input signals maximize accuracy multiple simple movements. We trained a Random Forest algorithm predict future angles across these movements using sensor features. hypothesized that angle would increase with addition IMUs attached adjacent body segments non-adjacent not accuracy. results indicated current inputs did significantly (RMSE 1.92° vs. 3.32° ankle, 8.78° 12.54° knee, 5.48° 9.67° hip). Additionally, including 5.35° 5.55° 20.29° 20.71° 14.86° 13.55° These demonstrated during improve alongside inputs.

Language: Английский

---

Use of a wearable electromyography armband to detect lift-lower tasks and classify hand loads

Applied Ergonomics, Journal Year: 2024, Volume and Issue: 119, P. 104285 - 104285

Published: May 25, 2024

Language: Английский

---

Classification Algorithms Trained on Simple (Symmetric) Lifting Data Perform Poorly in Predicting Hand Loads during Complex (Free-Dynamic) Lifting Tasks

Published: Jan. 1, 2024

Language: Английский

---

A Hierarchical-Based Learning Approach for Multi-Action Intent Recognition

Sensors, Journal Year: 2024, Volume and Issue: 24(23), P. 7857 - 7857

Published: Dec. 9, 2024

Recent applications of wearable inertial measurement units (IMUs) for predicting human movement have often entailed estimating action-level (e.g., walking, running, jumping) and joint-level ankle plantarflexion angle) motion. Although or information is frequently the focus intent prediction, contextual necessary a more thorough approach to recognition. Therefore, combination may offer comprehensive intent. In this study, we devised novel hierarchical-based method combining classification subsequent regression predict joint angles 100 ms into future. K-nearest neighbors (KNN), bidirectional long short-term memory (BiLSTM), temporal convolutional network (TCN) models were employed classification, random forest model trained on action-specific IMU data was used prediction. A action-generic multiple actions backward kneeling down, up, walking) also angle. Compared with approach, had lower prediction error up. TCN BiLSTM classifiers achieved accuracies 89.87% 89.30%, respectively, they did not surpass performance when in an model. This been because from actions. study demonstrates advantage leveraging large, disparate sources over Moreover, it efficacy IMU-driven, task-agnostic future across

Language: Английский

---

Classification algorithms trained on simple (symmetric) lifting data perform poorly in predicting hand loads during complex (free-dynamic) lifting tasks

Applied Ergonomics, Journal Year: 2024, Volume and Issue: 125, P. 104427 - 104427

Published: Dec. 10, 2024

Language: Английский

---