Capacitive Sensing for Natural Environment Biomechanics Monitoring

The included codebase illustrates how to use capacitive sensing data within two different wearable kinematics algorithms, CSInverseKinematics and CSOptimalControl. It shows how to load raw CS signals, process them, analyze them, learn from them, and predict kinematics with them on their own or in combination with other wearables.The included dataset comprises data from two experiments. The first dataset includes time-synchronized measurements of (1) muscle bulging acquired via a wearble lower limb capacitive sensing sleeve at the shank, (2) neural excitation measurements from electromyography, and (3) inferred muscle moments from static optimization performed in OpenSim with optical motion capture and instrumented treadmill data. 20 participants were recorded walking normally and with a 5-degree toe-in foot progression angle, a therapeutic modification used to mitigate progression of knee osteoarthritis. Measurements for CS and EMG were taken both inside a traditional motion capture laboratory environment and outside in natural environments.The second dataset includes measurements of (1) muscle bulging acquired via wearable lower limb capacitive sensing sleeves located at both the shank and thigh of both legs, (2) neural excitation measurements from electromyography, (3) optical motion capture and instrumented treadmill data, (4) XSens inertial measurement unit data, and (5) magnetic resonance imaging (MRI) body composition scan results. 10 healthy participants were recorded walking normally and with a mock impaired stiff-knee gait, along with 1 total knee arthroplasty patient. Measurements for CS, IMUs, and mocap were taking simultaneously, as well as measurements of EMG, IMUs, and mocap inside of the lab on an instrumented treadmill. The provided dataset enables the comparison of CS data with any biomarker in a consistent OpenSim/MATLAB ready formatting.Please cite the following when using this code or data: Owen Pearl, Nataliya Rokhmanova, Lauren Parola, Louis Dankovich, Summer Faille, Kenneth Urish, Sarah Bergbreiter, Eni Halilaj. (2024) Capacitive Sensing for Natural Environment Biomechanics Monitoring, Nature Communications (under review). https://doi.org/10.21203/rs.3.rs-1902381/v2. <br/><br/>This project includes the following software/data packages: <br/> <ul> <li> <a href="https://simtk.org/frs?group_id=2384#pack_2442">Code </a> </li> </ul>

本代码库阐述了如何在两种可穿戴运动学算法CSInverseKinematics和CSOptimalControl中运用电容感测数据。它展示了如何加载原始CS信号，对其进行处理、分析、学习，并利用其独立或与其他可穿戴设备结合的方式预测运动学。所包含的数据集由两个实验的数据组成。第一个数据集包含通过可穿戴下肢电容感测袖套在踝部获取的（1）肌肉膨胀时间同步测量数据，（2）从肌电图获得的神经兴奋测量数据，以及（3）在OpenSim中通过光学运动捕捉和仪器化跑台数据进行静态优化推断的肌肉力矩。20名参与者被记录以正常步态行走，并带有5度的内翻足进步角度，这是一种用于减轻膝关节骨关节炎进展的治疗性调整。CS和EMG的测量既在传统的运动捕捉实验室环境中进行，也在自然环境中进行。第二个数据集包括通过位于两条腿的踝部和股部的可穿戴下肢电容感测袖套获取的（1）肌肉膨胀测量数据，（2）从肌电图获得的神经兴奋测量数据，（3）光学运动捕捉和仪器化跑台数据，（4）XSens惯性测量单元数据，以及（5）磁共振成像（MRI）身体成分扫描结果。10名健康参与者被记录以正常步态行走，并带有模拟的僵硬膝关节步态，以及1名全膝关节置换术患者。对CS、IMUs和mocap的测量同时进行，同时在实验室内的仪器化跑台上对EMG、IMUs和mocap进行测量。提供的dataset允许在一致的开源Sim/MATLAB格式下将CS数据与任何生物标志物进行比较。在使用此代码或数据时，请引用以下文献：Owen Pearl, Nataliya Rokhmanova, Lauren Parola, Louis Dankovich, Summer Faille, Kenneth Urish, Sarah Bergbreiter, Eni Halilaj. (2024) Capacitive Sensing for Natural Environment Biomechanics Monitoring, Nature Communications (under review). https://doi.org/10.21203/rs.3.rs-1902381/v2. 本项目包含以下软件/数据包：
<ul>
<li><a href="https://simtk.org/frs?group_id=2384#pack_2442">代码</a></li>
</ul>