Body
Pattern recognition is a term frequently used and heard at the Center for Bionic Medicine. In short, pattern recognition describes using algorithms to decode and analyze a pattern (i.e., machine learning). Think, for example, of asking for Siri or Alexa on your phone or smart device. You have to first repeat the name several times before your device remembers the distinct pattern of your voice.
At the Center for Bionic Medicine, we focus on using pattern recognition algorithms specifically to decode muscle signals. Our researchers ultimately aim to use pattern recognition to develop prosthetic limbs that can be controlled more intuitively by the user. So rather than having the user initiate distinct muscle movements to 'command' the device, they can just think about how they want move and perform the desired motion.
Levi Hargrove, Director of the Center for Bionic Medicine, recently presented on this topic at an AbilityLab Research Seminary in July. Much of Dr. Hargrove's research investigates the potential of incorporating pattern recognition into motorized prosthetic limbs. His team's work on developing an Open Source Bionic Leg that incorporates pattern recognition was recently featured on Chicago Tonight with longtime research participant Terry Karpowicz.
"Every step, it learns a little bit more about Terry, and responds to his intentions a little bit better," Hargrove told Chicago Tonight. "So we're teaching the computer to respond to Terry or whoever else would use this leg, [and the prosthetic leg] would learn their habits a little bit better."
