Calibrated kinematic Ninapro hand movements data
收藏资源简介：
The kinematic data of the Ninapro (Non Invasive Adaptive Prosthetics) database includes calibrated kinematic data from 77 subjects, performing 40 hand movements and grasps. The whole database aims at allowing worldwide research groups to study hand kinematics. As well the dataset allows also to study the relationship between hand kinematics and muscle activity, since it is linked to sEMG datasets. The final goal of this work is to aid the progress in rehabilitation, physiotherapy, medicine, neuroscience and prosthetics. Therefore, the kinematic database improves the state of the art, being the most accurate, comprehensive and advanced reference for the largest kinematic database existing at the time of writing. Acquisition protocol: Hand kinematics was measured using a 22-sensor CyberGlove II dataglove (CyberGlove Systems LLC, www.cyberglovesystems.com). The CyberGlove is a motion capture data glove, instrumented with joint-angle measurements. It uses proprietary resistive bend-sensing technology to transform hand and finger motions into real-time digital joint-angle data. The calibrated kinematic Ninapro database includes 40 different movements of 77 intact subjects. The subjects have to repeat several movement represented by movies that are shown on the screen of a laptop. The experiment is divided in two exercises: • Isometric, isotonic hand configurations and basic wrist movements • Grasping and functional movements Joint angles from raw data were calculated according to a very detailed calibration protocol. The protocol was previously applied to 10 different intact subjects and consists in recording 64 different poses or guided movements to obtain the gains and also some corrections because of cross-coupling effects for specific anatomical angles. Gracia-Ibáñez, V., Vergara, M., Buffi, J. H., Murray, W. M. & Sancho-Bru, J. L. Across-subject calibration of an instrumented glove to measure hand movement for clinical purposes. C. Comput. Methods Biomech. Biomed. Eng. 20, 587–597 (2017). Data Sets: For each subject and exercise, the database contains one file in Matlab format (www.mathworks.com) with synchronized variables. The variables included are the following ones: • subject: subject number; • exercise: exercise number; • glove (22 columns): uncalibrated signal from the 22 sensors of the Cyberglove. Details on the location of the sensors are available at the link: ninapro.hevs.ch/node/123; • angles(22 columns): calibrated signal from the 22 sensors of the Cyberglove by applying an across-subject calibration • re-stimulus (1 column): the a-posteriori refined label of the movement; • re-repetition (1 column): re-stimulus repetition index; • stimulus (1 column): the label of the movement; • repetition (1 column): stimulus repetition index; • order of angles (22 columns) : name of the angles corresponding to variable “angles”. Raw glove data are also included in the folder "Raw data".
Ninapro（非侵入式自适应假肢，Non Invasive Adaptive Prosthetics）数据库的运动学数据包含来自77名受试者的标定后运动学数据，受试者完成了40种手部动作与抓握姿态。本完整数据库旨在支持全球研究团队开展手部运动学相关研究。此外，由于该数据集与表面肌电（sEMG）数据集关联，还可用于探究手部运动学与肌肉活动之间的关联。本数据集的最终目标是助力康复医学、物理治疗、临床医学、神经科学以及假肢技术领域的发展。因此，这款运动学数据库优化了当前技术水平，是截至本稿件撰写时已公开的规模最大、精度最高、内容最全面且技术最先进的运动学参考数据库。 采集方案：采用22传感器版本的CyberGlove II数据手套（CyberGlove Systems LLC，www.cyberglovesystems.com）采集手部运动学数据。该CyberGlove是一款运动捕捉数据手套，搭载关节角度测量模块，通过专有电阻式弯曲传感技术，将手部与手指的运动转换为实时数字化关节角度数据。经标定的Ninapro运动学数据库涵盖77名健康受试者的40种不同手部动作。受试者需重复播放于笔记本电脑屏幕上的影片所展示的多种动作。实验分为两项任务： • 等长、等张手部姿态与基础腕部运动 • 抓握与功能性动作 原始数据的关节角度计算遵循极为精细的标定方案。该方案此前已应用于10名健康受试者，流程包括记录64种不同姿态或引导式动作，以获取增益参数，并针对特定解剖角度的交叉耦合效应进行校正。 引用文献：Gracia-Ibáñez, V., Vergara, M., Buffi, J. H., Murray, W. M. & Sancho-Bru, J. L. Across-subject calibration of an instrumented glove to measure hand movement for clinical purposes. C. Comput. Methods Biomech. Biomed. Eng. 20, 587–597 (2017). 数据集详情：针对每名受试者与每项任务，数据库均包含一份Matlab格式（www.mathworks.com）文件，内含同步变量。所包含的变量如下： • subject：受试者编号； • exercise：任务编号； • glove（22列）：CyberGlove 22个传感器的未标定原始信号。传感器位置的详细信息可访问链接：ninapro.hevs.ch/node/123； • angles（22列）：通过跨受试者标定流程得到的CyberGlove 22个传感器的标定后信号； • re-stimulus（1列）：经后验优化的动作标签； • re-repetition（1列）：后验优化后的动作重复次数索引； • stimulus（1列）：原始动作标签； • repetition（1列）：原始动作重复次数索引； • order of angles（22列）：与"angles"变量对应的角度名称列表。 原始数据手套数据同样存储于"Raw data"文件夹中。
