This week Science Translational Medicine published an article outlining how researchers successfully gave amputees the ability to control their hands with their minds without undergoing an invasive or costly procedure; The key? Amplifying nerve signals via machine learning.
If given the ability to control their prosthetics with their mind, amputees could regain a wide variety of lost abilities which would, as a result, radically increase their standard of living.
Unfortunately, though this has been the subject of research for years, most solutions were incredibly invasive, as the peripheral nerves that could deliver the necessary signals from the spinal cord and brain are too tiny; to minimise this particular hurdle, scientists created an implant that amplifies these signals so that they may reach a prosthetic limb.
With the strength of these signals increased, researchers can then train algorithms to convert them to movements in real time. In fact, this process worked straight away for four patients.
They were able to “successfully control a hand prosthesis in real-time up to 300 days without control algorithm recalibration.”
Individual fingers could be controlled well enough to grasp small household objects and even play Rock, Paper, Scissors.
Though this study was conducted on amputees with prosthetic hands, it demonstrates the potential of such a process in giving those with other upper body amputations more control over their replacement limbs.