Estimation of Lower Extremity Joint Moments and 3D Ground Reaction Forces Using IMU Sensors in Multiple Walking Conditions: A Deep Learning Approach

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2023, Номер 27(6), С. 2829 - 2840

Опубликована: Март 29, 2023

Human kinetics, specifically joint moments and ground reaction forces (GRFs) can provide important clinical information be used to control assistive devices. Traditionally, collection of kinetics is mostly limited the lab environment because it relies on data that are measured from a motion capture system floor-embedded force plates calculate dynamics via musculoskeletal models. This spatially method makes extremely challenging measure outside laboratory in variety walking conditions due expensive device setup large space required. Recently, employing machine learning with IMU sensors suggested as an alternative for biomechanical analyses. Although these methods enable estimating human kinetic by linking sensor dataset, they show inaccurate estimates even highly repeatable single employment generic deep algorithms. Thus, this paper proposes novel model, Kinetics-FM-DLR-Ensemble-Net limb prediction hip, knee, ankle 3-dimensional GRFs using three thigh, shank, foot under several representatives daily living, such treadmill, level-ground, stair, ramp. first study implements both multiple learning. Our model versatile accurate identifying across diverse subjects outperforms state-of-the-art estimation margin.

Язык: Английский

---

Recent Advances in Wearable Healthcare Devices: From Material to Application

Bioengineering, Год журнала: 2024, Номер 11(4), С. 358 - 358

Опубликована: Апрель 6, 2024

In recent years, the proliferation of wearable healthcare devices has marked a revolutionary shift in personal health monitoring and management paradigm. These devices, ranging from fitness trackers to advanced biosensors, have not only made more accessible, but also transformed way individuals engage with their data. By continuously signs, physical-based biochemical-based such as heart rate blood glucose levels, technology offers insights into human health, enabling proactive rather than reactive approach healthcare. This towards personalized empowers knowledge tools make informed decisions about lifestyle medical care, potentially leading earlier detection issues tailored treatment plans. review presents fabrication methods flexible applications care. The potential challenges future prospectives are discussed.

Язык: Английский

---

A scoping review of portable sensing for out-of-lab anterior cruciate ligament injury prevention and rehabilitation

npj Digital Medicine, Год журнала: 2023, Номер 6(1)

Опубликована: Март 18, 2023

Abstract Anterior cruciate ligament (ACL) injury and ACL reconstruction (ACLR) surgery are common. Laboratory-based biomechanical assessment can evaluate risk rehabilitation progress after ACLR; however, lab-based measurements expensive inaccessible to most people. Portable sensors such as wearables cameras be deployed during sporting activities, in clinics, patient homes. Although many portable sensing approaches have demonstrated promising results various assessments related injury, they not yet been widely adopted tools for out-of-lab assessment. The purpose of this review is summarize research on applied ACLR offer our perspectives new opportunities future development. We identified 49 original articles ACL-related assessment; the common modalities were inertial measurement units, depth cameras, RGB cameras. studies combined with direct feature extraction, physics-based modeling, or machine learning estimate a range parameters (e.g., knee kinematics kinetics) jump-landing tasks, cutting, squats, gait. Many reviewed depict proof-of-concept methods potential clinical applications including screening, prevention training, By synthesizing these results, we describe important that exist validation existing approaches, using sophisticated modeling techniques, standardization data collection, creation large benchmark datasets. If successful, advances will enable widespread use portable-sensing identify factors, mitigate high-risk movements prior optimize paradigms.

Язык: Английский

---

Walking While Acting Sad and Happy Emotions Influences Risk Factors of Knee Osteoarthritis

Journal of Applied Biomechanics, Год журнала: 2025, Номер unknown, С. 1 - 8

Опубликована: Янв. 1, 2025

Greater knee adduction moment is associated with increased risk and progression of osteoarthritis, this biomechanical factor modulated through kinematic gait modifications. Emotions are known to influence walking kinematics speed, but the effect different emotions on mechanics unclear. To test this, 20 healthy participants walked while instrumented data was recorded. Participants initially naturally (baseline) then acting 4 emotional conditions: Anger , Happy Fear Sad in randomized order. Statistical parametric mapping an analysis variance model determined extent which influenced joint mechanics. Results indicated both happy ( P = .009) sad < .001) condition resulted lower compared baseline. Walking also speed changes from baseline .001). A secondary covariance as covariate no significant > .05), suggests that can be attributed speed. Decreased reduced osteoarthritis function, suggesting emotions, specifically sad, may moderate risk.

Язык: Английский

---

Analyzing Gait Angle Variations in Healthy Individuals and Knee Osteoarthritis Patients Utilizing Non-invasive IMU Sensors

Lecture notes in networks and systems, Год журнала: 2025, Номер unknown, С. 57 - 67

Опубликована: Янв. 1, 2025

Язык: Английский

---

Real-Time Ground Reaction Force and Knee Extension Moment Estimation During Drop Landings Via Modular LSTM Modeling and Wearable IMUs

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2023, Номер 27(7), С. 3222 - 3233

Опубликована: Апрель 27, 2023

This work investigates real-time estimation of vertical ground reaction force (vGRF) and external knee extension moment (KEM) during single- double-leg drop landings via wearable inertial measurement units (IMUs) machine learning. A real-time, modular LSTM model with four sub-deep neural networks was developed to estimate vGRF KEM. Sixteen subjects wore eight IMUs on the chest, waist, right left thighs, shanks, feet performed landing trials. Ground embedded plates an optical motion capture system were used for training evaluation. During single-leg landings, accuracy KEM R2 = 0.88 ± 0.12 0.84 0.14, respectively, 0.85 0.11 0.12, respectively. The best estimations optimal unit number (130) require placed selected locations landings. a leg only needs five leg's shank, thigh, foot. proposed LSTM-based optimally-configurable can accurately in relatively low computational cost tasks. investigation could potentially enable in-field, non-contact anterior cruciate ligament injury risk screening intervention programs.

Язык: Английский

---

Self-Supervised Learning Improves Accuracy and Data Efficiency for IMU-Based Ground Reaction Force Estimation

IEEE Transactions on Biomedical Engineering, Год журнала: 2024, Номер 71(7), С. 2095 - 2104

Опубликована: Фев. 5, 2024

Recent deep learning techniques hold promise to enable IMU-driven kinetic assessment; however, they require large extents of ground reaction force (GRF) data serve as labels for supervised model training. We thus propose using existing self-supervised (SSL) leverage IMU datasets pre-train models, which can improve the accuracy and efficiency IMU-based GRF estimation.

Язык: Английский

---

Deep Learning for Quantified Gait Analysis: A Systematic Literature Review

IEEE Access, Год журнала: 2024, Номер 12, С. 138932 - 138957

Опубликована: Янв. 1, 2024

Over the past few years, there has been notable advancement in field of Quantified Gait Analysis (QGA), thanks to machine learning techniques. QGA and gait prediction are areas where Deep (DL) techniques gaining popularity. There a significant amount attention from scientific community on application analysis various fields. Based our understanding, is noticeable absence comprehensive review current understanding utilizing DL Multi-task (MTL) models. Therefore, this paper provides assessment algorithms for QGA. The study takes systematic approach explore topic depth. We conducted thorough search three databases, namely Web Science, IEEEXplore, Scopus, identify relevant papers published 1989 October 2023. A total 55 were considered eligible included review. Approximately 46% studies that identified utilized classification models categorize phases locomotion modes. Additionally, portion (45%) regression estimate predict kinematic kinetic parameters, including joint angles, trajectories, moments, torques. Interestingly, 9% employed use MTL realm analysis. have also provided information most commonly datasets

Язык: Английский

---

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

Sensors, Год журнала: 2024, Номер 24(11), С. 3657 - 3657

Опубликована: Июнь 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.

Язык: Английский

---

Development of a Non-Contact Sensor System for Converting 2D Images into 3D Body Data: A Deep Learning Approach to Monitor Obesity and Body Shape in Individuals in Their 20s and 30s

Sensors, Год журнала: 2024, Номер 24(1), С. 270 - 270

Опубликована: Янв. 2, 2024

This study demonstrates how to generate a three-dimensional (3D) body model through small number of images and derive values similar the actual using generated 3D data. In this study, that can be used for type diagnosis was developed two full-body pictures front side taken with mobile phone. For data training, 400 datasets (male: 200, female: 200) provided by Size Korea were used, four models, i.e., recurrent reconstruction neural network, point cloud generative adversarial skinned multi-person linear model, pixel-aligned impact function high-resolution human digitization, used. The models proposed in analyzed compared. A total 10 men women analyzed, their corresponding verified comparing derived from 2D image inputs those obtained scanner. difference between an Unlike generation could not successfully various values, indicating implemented identify types monitor obesity future.

Язык: Английский

---