Multimodal hand movement dataset with EMG, IMU and finger kinematics in able-bodied and transradial amputees

Read full text Summarise with VIRAT
Authors: youssef nagaty abdelall ibrahim, Mohammed Ibrahim Awad., Nancy E. ElHady, Hend Abdelaziz Fayed, Ahmed M.Ebied
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The dataset comprises synchronized multimodal recordings of hand kinematics, muscle activity, and inertial signals acquired from both healthy participants and individuals with transradial amputation. A total of nine movement classes were defined to capture both isolated and coordinated finger actions. These include cyclic flexion and extension of each individual finger, combined movements of the index and middle fingers, coordinated motion of all fingers excluding the thumb, and two functional grasping tasks: a three-finger cup grip and a full-hand (five-finger) grip.All data were collected using the Delsys Trigno wireless sensing system and a Qualisys optical motion capture system. Surface electromyography (sEMG) signals were acquired from 14 channels using Delsys Trigno sensors at approximately 1259 Hz. Inertial measurements, including accelerometer and gyroscope data, were recorded from the same sensor units at 148 Hz. Hand kinematics were captured using the Qualisys motion capture system at 300 Hz, providing continuous three-dimensional marker trajectories and derived finger joint representations.All data streams were fully time-synchronized and recorded at their respective sampling frequencies. The dataset is organized into three primary data formats. The first format consists of JavaScript Object Notation (JSON) files containing the raw synchronized data, including 14-channel EMG signals, inertial measurements, and kinematic recordings. For each participant, the JSON files are organized into one folder per movement, and each folder contains up to three trials of that movement, named .json, where n indicates the order in which the trial was recorded. In each movement folder, the trial used for the analysis is marked with the suffix “export” (e.g., thumb_2_export.json); this is the trial from which the MATLAB and CSV files were generated. The number of trials available for each movement and participant is listed in Demographics_data.xlsx. The second format includes MATLAB (.mat) files containing the same data, where finger orientations are represented as joint angles in degrees rather than quaternions. The third format comprises comma-separated values (CSV) files, which provide a simplified representation of the dataset by including only the metacarpophalangeal (MCP) joint angles for each finger.This structured organization enables flexible use of the dataset for a wide range of applications, including signal processing, machine learning, and biomechanical analysis, while supporting both raw data access and simplified feature-level representations.

Fields of Study & Subjects

Muscle activation and electromyography studies Stroke Rehabilitation and Recovery Prosthetics and Rehabilitation Robotics

Publication details

Journal / source
ScienceDB
Publication date
23 March 2026
Language
EN
Document type
Dataset