Email
Password
Remember meForgot password?
    Log in with Twitter

article imageArtificial intelligence helps stroke patients walk again

By Tim Sandle     Jul 24, 2017 in Science
Lausanne - Researchers have developed an algorithm designed with a robot-assistive rehabilitation approach, to help people learn to walk again after neurological injuries such as stroke.
The new method to assist patients who have suffered severe neurological injuries is going through the clinical trial stage. The research comes from the Center for Neuroprosthetics and Brain Mind Institute, School of Life Sciences, Swiss Federal Institute of Technology Lausanne (EPFL), Lausanne, Switzerland. The regathers behind the development are hopeful it will lead to better outcomes for patients undergoing rehabilitation following incidences like a stroke or a spinal cord injury or strokes. With strokes, for example, many people experience weakness or stiffness in some of their muscles after a stroke. This happens because the brain sends signals to the muscles, through the nerves, to make them move. A stroke can damage the brain and affect these signals.
READ MORE: Mayo Clinic’s new startup to tackle diseases using AI
Recovery plans for spinal cord injuries and strokes typically require usually many hours of supported walking, using devices like treadmills, with the walking aid pre-programmed by a medic to provide a steady pace. This one-size-fit-all approach does not account for each patient moving around in different directions and having different gaits, both varying according to the individual.
With a new approach, a team led by Dr. Jean-Baptiste Mignardot have used digital technology to operate a robotic harness to help resist the downward force of gravity while also permitting patients to walk forwards, backwards, and side-to-side. This robotic harness is aided by an algorithm which can give personalized support to address patient-specific motor defects. In other words, recognizing that each patient is different.
SEE ALSO: Applying artificial intelligence to identify schizophrenia
The system is controlled by an artificial neural network that is capable of assessing by how much the upward and forward force needs to be varied, and to use this information to program the cable harness. The machine is able to assess 120 different variables relating to body movement and to apply the optimum requirements for the individual patient. This device has been tested, so far, on 26 patients (either recovering from spinal cord injuries or strokes). The participants were tested on four tasks—standing on two separate plates, walking on a straight path, walking on a wavy path, or walking on a ladder with irregularly positioned rungs. Each patient who took part in the first trial was able to walk with motor abilities comparable to healthy individuals.
The outcomes of the trial are shown in the following video:
It is hoped this practical framework, following further assessment, will be commercialized for use in the healthcare sector. The new development has been described in the journal Science Translational Medicine, with the research paper headed "A multidirectional gravity-assist algorithm that enhances locomotor control in patients with stroke or spinal cord injury."
More about Stroke, Artificial intelligence, Algorithm, Medicine
More news from