Problem statement and motivation

About 1500 babies each year are born with upper limb reduction in the US, including under-developed fingers, which is 4 out of 10,000 babies [3]. Missing limbs, including fingers, affect people’s daily life activities, such as:

  • Motor skills
  • Need assistance with daily activities such as self-care
  • Potential emotional and social issues because of physical appearance

Our patient:

  • Middle school girl
  • Right partial palm and partial thumb with finger nubs
  • Tested two 3-D printed body powered prosthetic
  • Would like a more affordable myoelectric prosthetic than what is commercially available

Our goal

Affordable Myoelectric Fingers is a project that aims to develop a solution for underdeveloped fingers via the design of a lightweight, affordable myoelectric finger that will be custom fit to a specific user. This includes:

  • Designing a custom specific, affordable, and lightweight myoelectric prosthetic fingers for a specific patient
  • Patient’s needs include
    • Index, middle, ring, little prosthetic fingers, and cosmetic thumb

Progress-to-date

In order to develop this prosthesis, the team is spending the semester investigating the physiology of the arm, the history and use of the previous prosthesis in order to improve on them as well as the function of the Myoband, an 8 channel sensor device that collects EMG signals from the arm.  The team has worked on designing an experiment to collect data from various team members to get EMG signal to be used for signal processing.  The goal for this semester is to develop a functional index finger that can extend and flex at various 45 and 90-degree angles.  Therefore the data is being collected at different angels.  After collecting the data the team worked on investigating filtering methods to remove noise from the data.  After this, we concluded that rather than filter the signal, which resulted in signal distortion, we should focus on using signal processing to determine at what angle the finger is flexing from the data.

The team used feature extraction to use specific features such as range, max, and min to differentiate between different angles.  Using this we have a  working prototype that can have finger flexion and extension at 45 and 90-degree angles.  Our prototype includes an Arduino and servo motor set up function in real time receiving the data processed from the myoband though a Matlab code we developed.  We will continue perfecting the prototype as well as making our prototype more compact.