The act of walking may not seem like a feat of agility, balance, strength and brainpower. But lose a leg, as Zac Vawter of Washington state did after a motorcycle accident in 2009, and you will appreciate the myriad calculations that go into putting one foot in front of the other. Taking on the challenge, a team of software and biomedical engineers, neuroscientists, surgeons and prosthetists has designed a prosthetic limb that can reproduce a full repertoire of ambulatory tricks by communicating with Vawter's brain. A report published on Wednesday in the New England Journal of Medicine describes how the team fit Vawter with a prosthetic leg that has learned - with the help of a computer and some electrodes - to read his intentions from a bundle of nerves that end above his missing knee. For those who have lost a leg or part of one due to injury or disease, Vawter and his robotic leg offer the hope that future prosthetics might return the feel of a natural gait, kicking a soccer ball or climbing into a car without hoisting an inert artificial limb into the vehicle. Vawter's prosthetic is a marvel of 21st-century engineering. But it is Vawter's ability to control the prosthetic with his thoughts that makes the latest case remarkable. If he wants his artificial toes to curl toward him or his artificial ankle to shift so he can walk down a ramp, all he has to do is imagine such movements. The work was done at the Rehabilitation Institute of Chicago under a US$8 million grant from the US Army. "We want to restore full capabilities" to people who have lost a lower limb, said Levi Hargrove, lead author of the new report. The report describes advances across a wide range of disciplines such as orthopaedic and peripheral nerve surgery. Weighing just over 4.5kg, the leg has two independent engines powering movement in the ankle and knee. It bristles with sensors, including an accelerometer and gyroscope, each capable of detecting and measuring movement in three dimensions. Most prosthetics in use today require the physical turn of a key to transition from one movement to another. But with the robotic leg, those transitions were effortless, Vawter said. "With this leg, it just flows," said the 32-year-old software engineer, who spends most of his days using a typical prosthetic but travels to Chicago several times a year from his home in Yelm, Washington. "The control system is very intuitive. There isn't anything special I have to do to make it work right." Before Vawter could strap on the bionic lower limb, engineers in Chicago had to "teach" the prosthetic how to read his motor intentions from tiny muscle contractions in his right thigh. At the institute's Centre for Bionic Medicine, Vawter spent many hours with his thigh wired up with electrodes, imagining making certain movements on command with his missing knee, ankle and foot. Using pattern-recognition software, engineers discerned, distilled and digitised those recorded electrical signals to catalogue an entire repertoire of movements. The prosthetic could thus be programmed to recognise the subtlest contraction of a muscle in Vawter's thigh as a specific motor command.