Design and evaluation of a bone-anchored, neurally-controlled knee prosthesis.

Single-participant case study with no formal hypothesis testing, so findings are descriptive and cannot be generalised to other people with transfemoral amputation. The participant had prior experience with neurally-controlled prostheses, which may make the observed adaptation and control better than first-time users would show. Powered actuation and neural control were tested together, so the paper cannot separate the contribution of either component. The neural knee was used only in supervised laboratory sessions; outside the laboratory the participant continued to use his microprocessor knee. The embodiment questionnaire was study-specific, not validated, and self-administered, so the higher scores are exploratory. Stair descent was not evaluated because the implant imposed a ±70 Nm torque limit, and stair ascent required handrails. Uninstructed standing was measured only in brief intervals before other tasks, so it may not represent a general standing strategy. Gait variability remained higher with the neural knee than with the prescribed device, and the authors state that gait patterns were still stabilizing after limited exposure.

The study was funded by the Massachusetts Institute of Technology (MIT). No other author conflict-of-interest declaration is provided in the supplied text.

Do not conclude that neural control alone improves mobility. The participant used a combined bone-anchored, powered, neurally-controlled system, and the paper says the contribution of powered actuation and neural control cannot be separated. This was one experienced user with no formal hypothesis testing.