A control system utilizing electrical signals was found to improve the real-time control of a powered prosthetic leg, according to results published recently in the Journal of the American Medical Association.
Levi J. Hargrove, PhD, an assistant professor of Physical Medicine and Rehabilitation at the Rehabilitation Institute of Chicago, and colleagues assessed the use of electromyographic (EMG) signal data from residual muscles with mechanical sensor data in a real-time control system on the ambulation performance of a powered prosthetic leg. Current powered-leg technology requires patients to slow down, stop, press buttons and perform various movements to transition between ambulation modes. Hargrove and colleagues wanted to determine whether EMG signals, which are regularly used to control powered-arm prostheses, could similarly provide real-time control for prosthetic powered legs.
According to a press release, the researchers studied seven patients with single-sided transfemoral or knee disarticulation amputations, all of whom were capable of ambulation using a passive prosthesis. Hargrove and colleagues conducted 20 trials of each patient, in which they studied the patients’ level-ground walking and stair and ramp ascent and descent. The researchers used pattern recognition algorithms to predict ambulation mode for the next stride using either mechanical sensor data alone or mechanical sensor data in combination with EMG data and historical information from earlier in the gait cycle.
The results showed the inclusion of EMG signals and historical information in the real-time control system resulted in a mean classification error of 7.9% across an average of 683 steps, a significant error reduction in comparison with the use of mechanical sensor data alone, which resulted in a classification error of 14.1% across an average of 692 steps.
The researchers noted the small sample size of the study and ambulation abilities of the patients are limitations requiring further research, according to the release.
Hargrove L, et al. JAMA. 2015;doi:10.1001/jama.2015.4527.
See the study for the full list of all authors’ relevant financial disclosures.