RoboDoc lends a hand to physios

FORGET RoboCop. British researchers are in the process of creating RoboDoc - or more exactly RoboPhysio - an "intelligent" machine capable of exercising patients whose limb movements have been impaired through stroke or head injury.

The research team, at the University of Leeds department of mechanical engineering, has made a prototype device that can "intelligently" flex the human elbow joint and is now working on a physiotherapy robot that can exercise the whole arm.

Researchers hope that a machine which responds to the patient's own movements will be a valuable tool for rehabilitation, enabling physiotherapists to supervise the treatment of several patients simultaneously.

The project was the idea of Dr Alastair Cozens, consultant in rehabilitation medicine at the Grampian Healthcare NHS Trust in Aberdeen. "One of the well-recognised problems for patients in neurological rehabilitation is that they do not receive the total amount of physical therapy that we would like them to receive," he said. "Anything that could increase this would be likely to be beneficial." A variety of machines already exists that simply move the patient's joints, but the key feature of a robotic physiotherapy device is for it to respond to the patient's own movements and to give assistance only when necessary.

For the prototype robotic elbow flexor, the patient's forearm is strapped to a motor-powered lever. Electronic sensors attached to the lever detect the angle of the elbow and the acceleration of the movement of the forearm and feed information to a computer, which communicates with the robot's motor.

The ideal trajectory of the patient's arm is fed into the computer by several "trial runs" with the physiotherapist guiding the patient's movements.

"The patient is presented with a semi-circular array of lights in front of their arm," said Dr Andrew Plummer, the project's supervisor. "When one of the lights is illuminated, they are required to move their hand to it. The robot's electronic sensors can detect if the patient needs assistance to complete the task."

Signals from the sensors are fed to a computer. If the patient cannot complete the task, the computer instructs the motor to assist the patient. "The key to the system is that the robot will assist the patient only if assistance is required. The patient must initiate any movement. In this way the robot acts much more like a real physiotherapist than simply as a passive exercise machine."

Research student Margaret Austin has been working with physiotherapists and stroke patients to fine-tune the control of the robot. "The work has shown that we can provide motor assistance without causing harm or unwanted muscle movement," she said. "The next stage is to try to develop a system that can assist with more complex exercises, such as reach-and- retrieve movements, where patients must extend their arms and pick up a cup, say."

"A robot system would not be intended to replace conventional physiotherapy, nor could it," said Dr Cozens. "The idea is that such robots could be used as an aid to physiotherapy, to provide additional exercise, or to enable a physiotherapist to supervise several patients simultaneously. The therapist would specify the treatment to be given by the machine, depending on the individual patient's requirements." The prototype the Leeds team is now working on, to allow the entire arm to be exercised, will have six motors operating two robotic arms which will work together to enable the patient's arm to move freely in all planes. It could be completed within a year.