3D printed prosthesis with CV, BCI and EMG

An affordable 3D printed prosthesis using computer vision to track and choose grip patterns for the user. Support BCI and EMG as well.

Similar projects worth following
This is a transhumeral prosthesis I have built to help a congenital limb deficiency patient in my country. The prosthesis is equipped with computer vision (object recognition and object tracking) to support wrist movement and choosing best grip patterns for object handling. The prosthesis is also equipped with a simple brain-computer interface and myoelectric sensor so that user could control the prosthesis at will, or give consent to the actions chosen by the prosthesis.
The prosthesis weights less than 1 kg, and has high dexterity. Overall, it has 10 DOF, controlled with Raspberry Pi and several Arduinos. Entire prosthesis is modular, so you could split the design to smaller prostheses (transradial prosthesis with wrist and hand, or just the hand).
I wrote several applications and software to go with the prosthesis, including an animated arm application that could be used to control the prosthesis (serial communication) and a logger application to record brain and muscle signals.

An update of the current state of the prosthesis. Demonstration of all the current supported features, including all mechanical movements, brain-computer interface and active vision system.

  • New EMG PCB finalized

    Nguyễn Phương Duy07/25/2019 at 04:44 0 comments

    To make the acquisition of EMG signals more easier and more convenient, as well as to increase the number of channels acquired while maintaining the portability of the prosthesis, here is a much smaller PCB design for EMG sensor:

    Current size is 30x16 mm.

  • EMG field test

    Nguyễn Phương Duy04/19/2019 at 03:43 0 comments

    The myoelectric sensor is field tested. The following videos and figures show the signal output and working of the sensor.

    The sensor PCB:

    The signal captured in oscilloscope:

    The working demo:

  • Active vision system in action with prosthesis

    Nguyễn Phương Duy04/19/2019 at 03:37 0 comments

    Putting the active vision system into the prosthesis. Demo videos:

    Vertical movements:

    Horizontal movements:

    Putting everything together:

    Rotated viewpoint for grip action:

    Here only the active vision is working, there is no consent from the user (BCI or EMG). When putting everything together, the system will be more robust and accurate.

  • BCI update with alpha trigger for hand grip action

    Nguyễn Phương Duy04/19/2019 at 03:32 0 comments

    First attempt of controlling the prosthesis with brain signal. The test is simple alpha trigger, using alpha signal to close and open the hand with a predefined grip pattern.

  • Fast active vision system developed for prosthesis

    Nguyễn Phương Duy04/19/2019 at 03:25 0 comments

    The following video shows the hybrid active vision system designed for this prosthesis. This system consists of an object tracker and an object recognition system that finds an object of interest, chooses the most appropriate grip pattern and waits for the user consent before attempting to grab the object. While the user moves closer to the object, the wrist will be calibrated.

    This is the first endeavor to create the real-time tracker for the next step.

  • Processing application to control prosthesis

    Nguyễn Phương Duy04/19/2019 at 03:20 0 comments

    An application is written to control the prosthesis from a PC. User could control the prosthesis by changing an animated arm on screen. The prosthesis will then be updated with new data through serial communication and calibrate itself.

    The application is written using Processing and Python.

    The purpose of this application is to either:

    • Support the user with another interface to debug the prosthesis, or practice using the prosthesis without damaging the mechanical parts (when practicing using EMG and brain-computer interface.)
    • Support the user with an interface to embed the prosthesis with more grip patterns in the future easily.

  • First design of the prosthesis

    Nguyễn Phương Duy04/19/2019 at 03:16 0 comments

    First design of the prosthesis with only 8 DOF. This design is frail and does not have high dexterity.

    This design does not have wireless support, everything is controlled using a built-in serial controller.

View all 7 project logs

Enjoy this project?



Jonathan Díaz wrote 07/18/2019 at 20:24 point

hi Nguyễn!

What a great project, congratulations for all the great work. We have a doubt do you hace any type of enconder or homing sensors for the fingers? 

  Are you sure? yes | no

Nguyễn Phương Duy wrote 07/19/2019 at 02:17 point

Hi Jonathan, there is a potentiometer at the shaft of each finger to feedback the position of the respective joint. The resolution is currently low, but it's usable somehow.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates