Today, the Motek Motion Technology team demonstrated the first results of the Computer Assisted Rehabilitation Environment (CAREN) before an audience of medical specialists and orthopaedists, who are partnering in the project. The invited physicians were offered a chance to test the motion platform's flexibility in a range of virtual tasks, designed for patients with balance disorder problems to help them regain their sense of coordination. As a result, the medical experts will be able to steer and advise the Motek team to adapt the virtual environments to the exact clinical needs of both doctors and patients. The system will be optimized by the end of the year and tests with real patients will follow in the course of 1999.
By means of a real time dataflow model with different scenarios, the CAREN tool-platform creates a feedback loop in which the individual actually forms part of the loop. The system thus registers dynamic motion behaviour, not only to assist patients suffering from balance disorder but also people with brain damage who fail to accurately coordinate the muscles because of lost motion memories. For the latter, the Motek team developed a tredmill with pre-registered feet data. While the patient is hanging with his feet on the platform, the tredmill reverses the walking process by simulating the motion of the feet.
Especially for the Centre of Orthopaedic Therapy (COT), the researchers have generated a scene to test the quality of prosthetics with flexible joints under extreme conditions. For the first time, it is possible to register the patient's progress visually by using overlay of data sets, captured at different times. A lot of the concepts applied for rehabilitation have been borrowed from the game industry. The game-play environment actually reduces the patient's fear and makes him feel safe. A good example constitutes the experiment with the staircase-effect. The psychological buffer, preventing the patient from walking, is removed because the platform floor is meeting his feet. Big surprise since the patient obviously needs less force to walk than imagined at first.
Today's demonstration set-up involved the installation of six cameras above the motion platform, a large projection screen, shutter glasses, and sensors to be worn on the person. The platform is steered by a Silicon Graphics dual-CPU OCTANE master server and a series of standard PCs. The Motek team tested four environments to give the audience an idea of the flexibility and reaction time of the connection between the operator and the platform. First, the platform was generated up/down and left/right via a computer mouse. Next, different states of calibration were shown, using full filters first for a safe and slow performance of balance correction by a person standing on the platform with two optical markers on the shoulders. The same scene without filters produced an immediate response time without any latency but here, the platform still acts a little jerky and calls for optimization.
The third environment takes the patient on a virtual boattrip on a sea with waves, regulated by the operator. The test person is wearing a magnetic sensor on the hips to simulate the roll of the boat. Again, the same scene is performed in two ways: with and without filters. To create a stereo vision effect, shutter glasses are used which offer a more realistic feeling of depth. In the fourth demonstration, a marker is placed on the centre of the platform in order to introduce sound into the simulation. If the marker is situated on the zero point, no sound can be heard but as soon as the patient deviates from the correct position, the sound is correlated to the degree of deviation.The medical professionals were invited to extensively test the environments and seemed delighted with the first results. Further input of the partnering medical experts is necessary to develop new suitable scenarios for swift integration into the system. Within five years, the Motek team plans to install the application in twenty specialized hospital sites all over Europe. The system can also be used in a wide range of industrial applications. For detailed information on the working of the CAREN tool-platform, we refer to the CAREN Web site and to the VMW report on CAREN in this very issue.