The Ninapro database: A resource for sEMG naturally controlled robotic hand prosthetics

Area of Science:

• Biomedical Engineering
• Rehabilitation Robotics
• Human-Computer Interaction

Background:

• Dexterous control of robotic prosthetics using non-invasive methods like surface electromyography (sEMG) presents significant challenges.
• Current sEMG-based control is often limited, unnatural, and requires extensive user training.
• Pattern recognition techniques show promise but require further development for practical application.

Purpose of the Study:

• To establish a benchmark database for developing and testing advanced algorithms for natural prosthetic hand control.
• To facilitate research in movement recognition and force control for robotic prosthetics.
• To improve the capabilities of naturally controlled robotic hand prosthetics.

Main Methods:

• The creation of the Ninapro database, a comprehensive collection of sEMG, kinematic, and dynamic data.
• Inclusion of data from 67 intact and 11 amputated subjects performing approximately 50 distinct movements.
• Development of a reproducible acquisition protocol for data collection.

Main Results:

• The Ninapro database provides a valuable resource for studying the relationships between sEMG signals, kinematics, and dynamics.
• The database enables the development and testing of sophisticated movement recognition and force control algorithms.
• Ongoing collaborations are expanding the database with new datasets.

Conclusions:

• The Ninapro database is crucial for advancing the field of naturally controlled robotic prosthetics.
• It supports the development of more intuitive and effective prosthetic hand control systems.
• Continued data expansion and research collaboration will accelerate progress in prosthetic technology.