Bionic prosthetic limbs: what is in use, what is in trials

Bionic prosthetic limbs are artificial arms and legs that use motors, sensors and control software to move. Some newer versions also connect to the wearer's bone, nerves or muscles. The aim is a limb that is easier to control and more comfortable to wear and, in some cases, one that can send back a sense of touch or position. Standard sockets do not suit everyone: the US approval record for one bone-anchored system lists socket problems such as skin infections and ulcers, pain and a short stump. This article separates three things: what regulators have approved, what has been tested in small studies, and what is still only reported.

This article is general information, not medical advice. If you have an amputation or are considering a prosthesis, speak to your limb fitting service or clinical team about your options.

How the main approaches differ

Most artificial limbs are held on by a socket and controlled by switches or by muscle signals read through the skin. Two newer approaches try to change that. In osseointegration, an implant is fixed into the bone and passes through the skin, so the limb attaches to the skeleton rather than to a socket. In nerve and muscle interfaces, surgeons redirect nerves or reconnect muscles so that the signals used to control the limb are stronger or easier to read. In some studies, electrodes also send sensory signals back to the nerves.

What regulators have approved

The approval we checked is for the OPRA Implant System, made by Integrum AB. The US Food and Drug Administration granted premarket approval (PMA P190009) with a decision date of 18 December 2020 (FDA approval record). The approval covers skeletally mature people with a transfemoral (above-knee) amputation due to trauma or cancer who have, or are expected to have, problems with a conventional socket prosthesis, or who cannot use one. The indication lists examples of those problems, including recurrent skin infections and ulcers in the socket contact area, pain, and a short stump. The record notes that the device has later supplements, so the indication may have changed since the original approval.

Approval means the device can be marketed for that group in the US. The sources we checked do not say how many clinics offer it or how many people use it, and we did not check its status with UK regulators, so we cannot describe it as routine care in the UK.

What has been shown in studies

Bone-anchored limbs. A single-centre, non-randomised study of 51 patients with 55 transfemoral amputations, enrolled between 1999 and 2007, reported 92% cumulative implant survival at two years. Superficial infection was the most frequent complication, with 41 episodes in 28 patients (Brånemark and colleagues, 2014). The five-year follow-up of the same group saw 40 of 43 eligible patients. Implant survival was still 92%, but only 45% of patients were free of any revision. Eleven patients had 14 deep infections, 15 patients had mechanical problems with the parts that connect the implant to the limb, and four implants were removed in total, one for infection and three for loosening. The authors wrote that "increases in deep infections and mechanical complications are concerning" (Brånemark and colleagues, 2018). The lead author of the five-year report declared employment, consultancy and shareholding links with Integrum, the maker of the implant.

Sensory feedback in very small groups. A 2014 study by Ortiz-Catalan and colleagues followed one person with an amputation for more than a year with an implanted, bone-anchored interface. Implanted electrodes gave more precise and reliable control than surface electrodes, and the team also reported sensory feedback produced by nerve stimulation (Ortiz-Catalan and colleagues, 2014). A 2014 study by Raspopovic and colleagues involved one person with a hand amputation. Stimulating nerves gave near-natural feedback, and the participant distinguished and consistently used three grip force levels without visual or auditory feedback (Raspopovic and colleagues, 2014). In 2007, Kuiken and colleagues reported one woman with an amputation near the shoulder, whose arm nerves were moved into chest muscles. Her score on a standard blocks test rose from 4.0 with her conventional prosthesis to 15.6 with the new one (Kuiken and colleagues, 2007). Each of these is a case of one person, so they show what is possible, not how common the result would be.

Walking after surgery. Song and colleagues, at MIT, studied 14 people with a below-knee amputation in a paper published in 2024. Seven had a surgery called the agonist-antagonist myoneural interface (AMI), which reconnects pairs of muscles that normally work against each other, and seven were matched controls with a standard amputation. In the surgery group, signals from the reconnected muscles reached about 18% of the level found in an intact limb. Maximum walking speed was 1.78 m/s against 1.26 m/s in the control group, 41% higher. The surgery group also adapted to slopes, stairs and an obstacle while the limb was under continuous nerve-signal control (Song and colleagues, 2024). Participants had two practice sessions, about six hours of walking in total, and we found no long-term follow-up reported. The study was non-randomised, and participants and researchers were not blinded. The authors state that future studies should show clinical benefit from long-term use. Two authors are inventors on patents covering the procedure.

Reported, not yet shown

• Arm amputations. The 2024 paper says the AMI procedure has been performed at arm amputation levels as well as leg levels, and MIT's press release says it can be done for people with arm amputations (MIT News, 2024). The 2024 study tested leg amputations only, and we did not find published results for neural control of arm prostheses after this surgery.
• Patient numbers. The 2024 paper says the procedure has been performed on more than 50 patients, and the MIT press release says about 60 worldwide. We did not find an independent count to check either figure.

Limits and open problems

• Infection and skin. The implant passes through the skin, so infection is a central concern. Superficial infection was the most frequent complication in the two-year results. The five-year follow-up reported 14 deep infections in 11 patients. The authors also reported that mechanical complications were more frequent in more active patients.
• Fitting and training. The sources we read give no figures for fitting time or overall training time. A 2025 German-language review of the AMI surgery says it remains to be re-evaluated whether the extra operative and time burden is justified by the functional gains, and for which patient groups the method is suitable (Kalff and colleagues, 2025).
• Battery and weight. In the 2024 study, the battery module ran the bionic system for about two hours of continuous use per charge, and the ankle-foot unit weighed 2.42 kg without the battery. We found no data on repair or replacement of these devices.
• Cost and access. The FDA record and the studies we read give no costs. The 2025 review contains no cost or reimbursement data.
• Evidence. The sources we read include no randomised trial comparing these limbs with standard prostheses. The 2025 review states that the authors' work was supported by research funding from Ottobock SE & Co. KGaA, a prosthetics maker.
• Lab to daily life. The 2024 tests covered a set of walking tasks in controlled conditions. Sprinting, jumping and single-leg balancing were not tested, and the authors say future testing might include them.

What to watch next

• A randomised or blinded comparison with standard prostheses.
• Implant survival and infection figures beyond five years for the approved system. The FDA record links to a post-approval study, but we have not reviewed its design or status.
• Published results for nerve and muscle interfaces in people with arm amputations.
• Cost, reimbursement and access decisions for the surgery and the prosthesis.
• Longer battery life and lower weight in later nerve-controlled prostheses.