Skip to content

A Final Project Program, Design and Build of Medical Rehabilitation Device for People with Paralysis on Hand Based on EMG Sensor

Notifications You must be signed in to change notification settings

muhardianab/EMG-Rehab-Device

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

55 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

EMG-Rehab-Device

This repository is about my Final Project Program, Design and Build of Medical Rehabilitation Device for People with Paralysis on Hand Based on EMG Sensor

Final Project Implementation

This device help people with paralysis on hand to strength his/her hand and as a rehabilitation to recover his/her hand back to normal and healthy

A person with a disability cannot do work freely or is limited, one of which is paralysis of the hand caused by a stroke or spinal injury. Hands as a means of movement that are often used by humans play an important role in their daily lives, therefore human dependence on their body tools is very large. To be able to cure paralysis, one of them is by making a rehabilitation tool to help someone return the condition of his paralyzed hands to normal. A rehabilitation tool that is made to stimulate the movement of the muscles in the fingers and help them move stronger in closing and opening the hand. An exoskeleton-shaped device based on an Electromyogram (EMG) sensor can support human hands as well as autonomous rehabilitation tools that can be moved without external assistance. The exoskeleton itself is in the form of a glove equipped with a manipulator made of solid material that can support the hands. The system used in the tool is a pattern recognition system to be able to classify basic movements of opening and closing hands and variations in grip strength. The classifications used are Back Propagation Artificial Neural Network (BPANN) and linear regression with 720 training data and 180 testing data. The resulting accuracy is 83.89% in offline testing. For online testing itself, it produces 74% accuracy with 50 trials in 5 classes.

Datasets

There are 2 datasets based on the code are used to take the data,

  1. Dataset 1 is using code from https://github.com/mark-toma/MyoMex
  2. Dataset 2 is using code from https://github.com/Lif3line/Myo-MATLAB-Interface-EMG-GYRO-ACCEL

Methods

There are 8 sensors or channels on Myo Armband and each sensor is extracted into 8 features. Those features were trained using NN from https://github.com/vtshitoyan/simpleNN

Documentation

Here are the video I've documented

If you want to know more detail about my final project, just email me to [email protected]

About

A Final Project Program, Design and Build of Medical Rehabilitation Device for People with Paralysis on Hand Based on EMG Sensor

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published