Related Work:
Cedeño et Al. proposed the use of Analysis of Electromyographic signals (EMG) from wearable called MYO and extracted only one feature, the Root Mean Square (RMS) obtaining accuracy close to 95%, but the system based on FPGA needs 2 seconds of data recorded and is a post-acquisition system.
Some authors used the EEG signals to control the finger prosthesis movements using the low cost ”Emotiv” device, which comprises 14 channels.
Clustering of EEG Occipital Signals using K-means, were author use Steady state visual evoked potential (SSVEP), the authors used occipital area of scalp [25].
BCI2000 instrumentation system.
Implementation of Neural Network-Based on EEG signals Classification Approach on FPGA, but is a post-acquisition system and motor cortex activity [26].
Discussion and Conclusions:
Our solution is:
Simple and low cost 3D printed prosthetic forearm
Right placement of the prosthesis on the stump
We get Provide comfort to the amputee
Driven by EEG signals and powered by three servo motors
Used a single sensor placed on the forehead
Use biodegradable materials for its manufacture as PLA
We find the proper EEG data (attention value) and transfer it wirelessly to the electronic control devices that send the control signals to the servo actuators according to the subject's movement intention.
Simple control algorithm based on a range of attention has been implemented
Allowing the patient to open or close the hand according to the concentration level.
This prosthesis could be useful for people who require a lightweight and straightforward prosthesis, low cost, and binary operation in order to be able to grasp and release objects used in daily life.
Space is optimized in the forearm’s upper part to contain the electronic components, including the two batterys (800mA).
Our solution is:
Simple and low cost 3D printed prosthetic forearm
Right placement of the prosthesis on the stump
We get Provide comfort to the amputee
Driven by EEG signals and powered by three servo motors
Used a single sensor placed on the forehead
Use biodegradable materials for its manufacture as PLA
We find the proper EEG data (attention value) and transfer it wirelessly to the electronic control devices that send the control signals to the servo actuators according to the subject's movement intention.
Simple control algorithm based on a range of attention has been implemented
Allowing the patient to open or close the hand according to the concentration level.
This prosthesis could be useful for people who require a lightweight and straightforward prosthesis, low cost, and binary operation in order to be able to grasp and release objects used in daily life.
Space is optimized in the forearm’s upper part to contain the electronic components, including the two batterys (800mA).
⭐ For more information visit our blog:
https://vasanza.blogspot.com/
1. A 3D-Printed EEG based Prosthetic Arm
Josue Fuentes Gonzalez, Andres Infante Alarcón,
Víctor Asanza and Francis Loayza
Escuela Superior Politécnica del Litoral, ESPOL, Guayaquil, Ecuador
Facultad de Ingeniería en Electricidad y Computación, FIEC
Facultad de Ingeniería en Mecánica y Ciencias de la Producción, FIMCP
2. • Introduction
• Related Work
• Methodology
• Results
• Discussion and Conclusions
Topics
A 3D-Printed EEG based Prosthetic Arm
4. • Cedeño et Al. proposed the use of Analysis of Electromyographic signals (EMG) from wearable called MYO
and extracted only one feature, the Root Mean Square (RMS) obtaining accuracy close to 95%, but the
system based on FPGA needs 2 seconds of data recorded and is a post-acquisition system.
• Some authors used the EEG signals to control the finger prosthesis movements using the low cost ”Emotiv”
device, which comprises 14 channels.
• Clustering of EEG Occipital Signals using K-means, were author use Steady state visual evoked potential
(SSVEP), the authors used occipital area of scalp [25].
• BCI2000 instrumentation system.
• Implementation of Neural Network-Based on EEG signals Classification Approach on FPGA, but is a post-
acquisition system and motor cortex activity [26].
Related Work
BCI2000 instrumentation system
MYO EMOTIV EPOC
5. Location of the Neurosky sensor in Frontopolar region one
and virtual ground in the left ear in the 10-20 System EEG
Methodology
Neurosky sensor
7. 64 surface EEG Electrodes International System 10-10
Methodology
EEG Signals
Data Acquisition
Features Extraction Activación de la prótesis
ON ON ON ON
OFF OFF OFF
ON: Activation of the prosthesis (closed hand) when the attention value is
greater than 60. OFF: Deactivation of the photesis when the attention value is
less than 60 (open hand)
Attention Value
60
8. • EEG Signals Data Set
• 5 Healthy Control Subjects
• 5 Amputee Subjects
Methodology
EEG Signals
Data Acquisition
Features Extraction Activación de la prótesis
9. Methodology
The left side shows the prosthesis design made with Blender. The
right side shows the 3D printed device, including the servomotors.
MG995 servomotor’s torque is 8.5 N per cm, with
operating voltage of 5 V DC.
11. General connection diagram. A. Neurosky device over the head. B. Bluetooth module HC-05. C. 5 [V] batteries. D.
Arduino Micro. E. Servomotors. F. Draft Hand.
Methodology
13. Transmission force to the fingers through the nylon thread, the
servomotor emits a specific torque of 8.5 Kgf per cm
Results
• The measured value in each finger was 2.22 N. on average by finger. Finally, to determine the total force of the hand,
multiply by five as follows in equation (1).
• This means that the total grip strength of the 3D printed prosthesis was approximately 11.0 N. So, we need to use
three servo motors.
14. Discussion and Conclusions
• Our solution is:
• Simple and low cost 3D printed prosthetic forearm
• Right placement of the prosthesis on the stump
• We get Provide comfort to the amputee
• Driven by EEG signals and powered by three servo motors
• Used a single sensor placed on the forehead
• Use biodegradable materials for its manufacture as PLA
• We find the proper EEG data (attention value) and transfer it wirelessly to the electronic control devices that
send the control signals to the servo actuators according to the subject's movement intention.
• Simple control algorithm based on a range of attention has been implemented
• Allowing the patient to open or close the hand according to the concentration level.
• This prosthesis could be useful for people who require a lightweight and straightforward prosthesis, low cost,
and binary operation in order to be able to grasp and release objects used in daily life.
• Space is optimized in the forearm’s upper part to contain the electronic components, including the two batterys
(800mA).
15. • This functional prosthesis’s utility based on EEG signals is suitable for upper limb amputee people, especially
for those who cannot obtain a reliable EMG signal due to the injury severity.
• Other prosthesis use the placement of various superficial muscle sensors over the skin or internally
implanted in the muscle.
• Future Works:
Diagram of the pattern recognition function of neural networks in Simulink
Multi-Layer Perceptron (MLP)
Discussion and Conclusions
Open BCI headset
16. Cite This
• Published in
• 2020 IEEE International Conference on E-health Networking, Application & Services
(HEALTHCOM)
• DOI
• 10.1109/HEALTHCOM49281.2021.9399035
• Conference Location
• Shenzhen, China
• Plain Text
• J. Fuentes-Gonzalez, A. Infante-Alarcón, V. Asanza and F. R. Loayza, "A 3D-Printed EEG
based Prosthetic Arm," 2020 IEEE International Conference on E-health Networking,
Application & Services (HEALTHCOM), 2021, pp. 1-5, doi:
10.1109/HEALTHCOM49281.2021.9399035.
17. For more information
Víctor Asanza
Mail: vasanza@espol.edu.ec
Facultad de Ingeniería en Electricidad y Computación, FIEC
Escuela Superior Politécnica del Litoral, ESPOL
Campus Gustavo Galindo Km 30.5 Vía Perimetral, P.O. Box 09-01-5863
090150 Guayaquil, Ecuador
Personal Blog: Orcid: ResearchGate: Google Schoolar: