Experiences in the Creation of an Electromyography Database to Help Hand Amputated Persons
0 likes2,261 views
This document discusses the creation of an electromyography (EMG) database to help hand amputated persons. The goals were to develop an acquisition protocol, acquire a public database, and test classification algorithms using a common dataset. Sensors were placed on arm muscles and a data glove recorded hand movements during exercises. The database includes clinical data and unprocessed signals from 27 non-amputated and 1 amputated subject. The acquisition protocol and sensors were tested and refined. The database is available online and aims to improve prosthesis control through improved EMG classification.
Experiences in the Creation of an Electromyography Database to Help Hand Amputated Persons
- 1. Experiences in the Creation of an Electromyography Database to Help Hand Amputated Persons Manfredo Atzori, Arjan Gijsberts, Simone Heynen, Anne-Gabrielle Mittaz Hager, Claudio Castellini, Barbara Caputo, Henning Müller
- 2. Overview • EMG and prosthetics • Motivations and goals • Acquisition setup and sensors • Hand movements • Results – Electrodes 2
- 3. Electromyography • Electromyography (EMG) is the measurement of electrical activity that is created by muscle contractions • The signal path – Originates in a motor neuron – Travels to the target muscle(s) – Starts a series of electrochemical changes that lead to an action potential – Can be detected by one or more electrodes 3
- 4. EMG controlled prosthetics • 2-3 degrees of freedom • Few programmed movements • Very coarse force control • No dexterous control • No natural control • Long training times • This is in contrast to recent advances in mechatronics! 4
- 5. Motivation for the work • NO large scale public sEMG databases, only private ones – (Fukuda, 2003; Tsuji 1993; Ferguson, 2002; Zecca, 2002; Chan, 2005; Sebelius, 2005; Castellini, 2008; Jiang, 2009; Tenore, 2009; Castellini, 2009) • NO common sEMG acquisition protocol • NO common sEMG storage protocol • NO Clinical Data Correlation Evaluation 5
- 6. Goals • Creation and refinement of an acquisition protocol • Acquisition of a database • Public release of the database • Worldwide test of classification algorithms using the same data and setup – Improve quality of classification – Transfer this knowledge to build better prostheses 6
- 7. Acquisition setup • Laptop: Dell Latitude E5520 • Digital Acquisition Card: National Instruments 6023E • sEMG Electrodes: 10 double-differential Otto Bock 13E200 • Printed Circuit Board, Cables & Connectors • Data Glove 22 sensors Cyberglove II (Cyberglove Systems) • Inclinometer: Kübler 8.IS40.2341 7
- 8. Sensor setup • 8 equally spaced electrodes • 2 electrodes on finger flexor and extensor muscles • Two axes inclinometer • Data glove 8
- 9. Hand movements Training Hato, 2004 Exercise 1 Sebelius, 2005 12 movements Farrel, 2008 Crawford, 2005 Exercise 2 17 movements 9
- 10. More movements Exercise 3 23 movements Feix, 2008 DASH Score Objects are simple tools to make the protocol easy to reproduce everywhere. 10
- 12. Data stored • XML file with clinical and experimental information • Unprocessed data (sEMG, Cyberglove, Inclinometer, Movie) • Preview picture for each exercise • Picture of the arm without the acquisition setup • Picture of the arm with the acquisition setup • Currently stored: 27 intact subjects, 1 amputated subject, several recordings for a few Gender 21 males 7 females Handedness 26 right handed 2 left handed Age 28.1 ± 3.4 years 12
- 13. Electrode experiences • Double differential potential – Good signal to noise ratio • Excellent comfort (no cleaning/ shaving) • Classification results in accordance with the scientific literature (~7-20%) 13
- 14. Two types of electrodes tested • Otto Bock 13E200 – Root mean square rectification – High pass filtering – Sampling frequency: 100 Hz L. F. Law et al., 2010 • Delsys Trigno – Raw signal – Sampling frequency: 2KHz – Wi-fi L. F. Law et al., 2010 14
- 15. Acquisition experiences (amputated) • Dry the stump before the experiment • Need of longer breaks between the exercises • Modification of the instructions avoiding the concept of an imaginary limb • Elimination of a few movements from the protocol 15
- 16. Acquisition experience (non amputated) • Difficulty to place electrodes exactly in the same position for subjects – Need of spatial normalization as anatomy changes and positions are not 100% stable • Validation of the acquisition protocol with small changes – Function check of electrodes is required – System needs to limit artifacts caused by users • Removal of a few functional movements showing high inter subject differences 16
- 17. Conclusions • Test and improvement of the acquisition setup – Portable, based on research and industrial needs • Test and improvement of the acquisition protocol – Complete and easy to be reproduced – Fixed several practical aspects • Test and improvement of the hand movements – 52, selected from robotics and medical literature • Test of acquired sEMG signals (classification) – Good SNR ratio – Results in line with the scientific literature 17
- 18. Questions, contacts? • For more information: • http://www.idiap.ch/project/ninapro/ • http://ninapro.hevs.ch • Contacts: • manfredo.atzori@gmail.com • henning.mueller@hevs.ch 18