EMG elbow dataset

This dataset contains the elbow's surface electromyography (sEMG) and kinematic (joint angle) data of ten subjects (6 males and 4 females) when performing different exercises of Flexion - Extension and Pronation - Supination. For the Flexion - Extension movements, the data was collected with the subject standing performing the movement with his arm next to the body; for the Pronation - Supination movements, the subject was seated in a chair with his forearm supported on a table. The EMG data was acquired using a Shimmer3 EMG device and the joint angle was estimated using the integrated IMUs in the Shimmer3 device. The following excercises were performed by each subject: Flexion - Extension without load, Flexion - Extension with 3lb dumbbel, Flexion - Extension with 5lb dumbbell, Pronation - Supination without load, Pronation - Supination with 3lb dumbbell and Pronation - Supination with 5lb dumbbell. The data is stored in a folder for each one of the subjects, each folder contains the following 13 files: one file named subject_info includes the subject ID, age, height, weight, arm length, hand width and gender. The remaining 12 files (2 for each exercise) are named with the following convention <id_movement_set_load[g]> (id: subject id, movement: excercised performed, set: training or test set, load: external load in grams). The data in each file is structured as follows: raw sEMG Channel 1 | raw sEMG Channel 2 | filtered sEMG Channel 1 | filtered sEMG Channel 2 | joint angle For Flexion - Extension: Channel 1 - Biceps Brachii and Channel 2 - Triceps Brachii. For Pronation - Supination: Channel 1 - Pronator Teres and Channel 2 - Biceps Brachii. The sEMG signal was filtered using a fourth order Chebyshev filter with cutoff frequencies of 10Hz and 500Hz, using a band ripple of 5%.

本数据集收录10名受试者（6男4女）在完成屈伸运动（Flexion-Extension）与旋前-旋后运动（Pronation-Supination）两类不同动作时的肘部表面肌电（surface electromyography，sEMG）及运动学（关节角度）数据。针对屈伸运动，数据采集于受试者站立、手臂紧贴躯干完成动作的场景；针对旋前-旋后运动，数据采集于受试者坐姿、前臂支撑于桌面的场景。肌电数据通过Shimmer3肌电设备采集，关节角度则通过该设备内置的惯性测量单元（Inertial Measurement Unit，IMU）估算得到。每名受试者需完成以下6项运动：无负重屈伸运动、佩戴3磅哑铃的屈伸运动、佩戴5磅哑铃的屈伸运动、无负重旋前-旋后运动、佩戴3磅哑铃的旋前-旋后运动、佩戴5磅哑铃的旋前-旋后运动。数据按受试者分文件夹存储，每个受试者文件夹包含13个文件：其中一个名为subject_info的文件记录受试者编号、年龄、身高、体重、臂长、掌宽及性别；剩余12个文件（每项运动对应2个文件）遵循以下命名规则：<id_movement_set_load[g]>，其中id为受试者编号，movement为所完成的运动类型，set为训练集或测试集，load为以克为单位的外部负重。每个文件内的数据结构如下：原始sEMG通道1 | 原始sEMG通道2 | 滤波后sEMG通道1 | 滤波后sEMG通道2 | 关节角度。针对屈伸运动，通道1对应肱二头肌，通道2对应肱三头肌；针对旋前-旋后运动，通道1对应旋前圆肌，通道2对应肱二头肌。表面肌电信号采用四阶切比雪夫滤波器进行滤波处理，截止频率为10Hz与500Hz，带纹波为5%。

搜集汇总

背景与挑战

背景概述

该数据集收集了10名受试者在不同负重条件下进行肘部屈伸和旋前旋后运动时的表面肌电信号(sEMG)和关节角度数据。数据包含原始和滤波后的双通道sEMG信号，并按受试者文件夹和运动类型分类存储，适用于研究肌肉活动与关节运动的关系。