Sensational

Researchers at the University of Utah have begun a four-year project to develop a bionic arm that moves naturally in response to thought and is capable of feeling.

The key to the research will be the development and testing of a 'peripheral nerve interface' — an implanted device that will relay impulses from nerves in the residual limb (the portion of limb remaining after amputation) to a small computer worn on a belt and then the bionic arm.

Sensors in the arm will send signals to nerves in the remainder of the amputated arm and then to the brain, allowing the user to sense the arm's motion and location, and to feel objects with the mechanical hand and fingers, invoking perception and feeling.

Key to this interface will be the Utah Slanted Electrode Array — a wireless device that will be able to interpret signals from residual nerves, translate them and send them wirelessly to the bionic arm. Greg Clark, principal investigator at

, explained how the device, no bigger than 4mm2 can talk to axons (primary transmission lines of the nervous system) in the arm. 'The array contains 100 individual electrodes that can talk selectively to the biological wires in the nerve, so that if a user thinks of making a movement, the signal goes from the brain, via the spinal cord to the muscle.

'At the same time, we can have sensors on fingertips or to measure the joint angle to establish sensory, perceptive information and send it back the other way.'

This technique differs from conventional prosthetics in two ways. traditional prosthetics are controlled by signals from an intact muscle, for example a shoulder shrug which must be learned by the user and can be challenging.

Second, only one movement can occur at a time, whereas the new prosthesis will tap into signals going to all the muscle groups. This means the arm could merge various movements simultaneously to produce a more realistic movement. As

said, it is a way to listen in to what the nervous system would be telling the natural arm and translating that into signals that will move the artificial arm the same way.

Despite being involved in this field for the past 15 years,

admitted that the project is an ambitious and challenging one. 'One of the challenges is making the array itself,' he said. 'To send signals wirelessly from the electrode to the arm requires a lot of integrated circuitry on the back of a small device. You need 100 low-noise, low-power amplifiers that can take the neural signals from each of the electrodes, some digital signal processing, some telemetry to send the data out — and of course it has to be powered — before being packaged in the 4mm2 device.'

Worth potentially $10.3m (£5.5m) the project is just part of a larger DARPA-funded initiative called Revolutionising Prosthetics which involves the John Hopkins University Applied Physics Laboratory and a number of sub-contractors, including two branches of the Fraunhofer Institute; the Institute of Reliability and Microintegration and Institute for Biomedical Technology who will assist in the manufacture of the array and components.

The Pentagon announced the project earlier in the year when it awarded John Hopkins $30.4m over two years, with additional funding that could bring the total to $54.8m over four years. The university is sub-contracting work to 28 other universities, labs, hospitals and companies who will be involved in making the rest of the arm, such as sensors, power supply and attachments.

But whether, after just four years, this project will have produced a fully-fledged realistic prosthetic limb remains to be seen.

, however, is confident that the project will achieve a significant step forward for those with lost limbs to recover a normal life. 'Yes, I think we can make major advances in control and sensory feedback to the system. This project will make an extraordinary difference between what exists today and what will exist in a few years' time,' he said.