Pattern Recognition Technology for Upper Limb Prosthesis Users

Welcome to our new blog, Dankmeyer Download!!  Hosted by Nina Bondre, CPO, each month we will introduce a topic of interest or recent research in our field.  We hope you enjoy these discussions of the latest research, techniques, and developments in the fields of orthotics and prosthetics.

Dankmeyer, Inc. and Infinite Biomedical Technologies (IBT) hosted a course together at Dankmeyer’s Baltimore Washington area office (Linthicum, MD) in early April, focusing on new technologies for upper limb prosthetic users. Sean McHugh from IBT presented to a group of occupational therapists from Johns Hopkins Physical Medicine and Rehabilitation and practitioners from Dankmeyer about IBT’s new pattern recognition technology, Sense. We were fortunate to have several of our patient models attend to allow participants to work with upper limb prosthesis users. Participants had the opportunity to find ideal electrode locations on patient models' residual limbs, and also try out the IBT Sense training and programming software. 

Myoelectric prostheses are prostheses that use the EMG (electromyography) signals from muscle contractions to drive actions. Traditionally, prosthetists would work with patients to identify distinct muscle contractions and signals, and place the electrodes accordingly. Electrodes pick up the electrical muscle signals. Based on what signal the user generates, either by contracting one muscle group versus another (and in some cases, based on length of time and contraction strength), different actions occur in the prosthesis. For example, one muscle group/electrode may be associated with opening the prosthetic hand, and the other muscle group/electrode will close the prosthetic hand. 

Pattern recognition still uses EMG signals from electrodes, but operates off of the combination of signals from these various electrodes. Whereas prosthetists would need to carefully locate electrodes in the previous designs and make sure they were not overlapping, pattern recognition focuses on using the distinct patterns generated by someone's muscle contractions, rather than firing one muscle group versus another. This allows for more intuitive control of the prosthesis, and allows this type of prosthesis to be available to a wider range of users. Traditional myoelectric control for some users was challenging, as the requirements for distinct muscle signals controlled so carefully was difficult to train. One of the advantages of the IBT system is that there is intuitive training software that allows users to train on the system and fine tune their patterns at home. 

You can read more about IBT's Sense system by clicking here: Sense system. We recently featured IBT and some of this research in a patient story, as told by Dr. Andrew Rubin. You can read his story by clicking here.

If you have any questions for Nina about this topic or any other, email us at Suggestions are also welcome!

Here is a gallery of photos from the course!