First successful non-invasive mind-controlled robotic arm.
Robotic instruments have been assisting clinical staff and engineers alike for many years, with prosthetic robotic arms, which allow people who have lost a limb to regain freedom of movement, currently in development. Brain-computer interface (BCI), a collaboration between the brain and a device, enables signals from the brain to direct some external activity, such as control of a cursor or a prosthetic limb has garnered much interest in robotic prosthetics. However, to-date BCIs successful in controlling robotic arms have needed highly invasive brain implants and surgeries. Now, a study led by researchers at Carnegie Mellon University develops a non-invasive, mind-controlled robotic arm which can continuously track and follow a computer cursor. The team state being able to non-invasively control robotic devices using only thoughts will have broad applications, in particular benefiting the lives of paralyzed patients and those with movement disorders. The opensource study is published in the journal Science Robotics.
Previous studies show a grand challenge in BCI research is to develop less
invasive or even totally non-invasive technology which would allow paralyzed patients to control their environment or robotic limbs using only their own thoughts. Such non-invasive BCI technology would bring such much needed technology to numerous patients and the general population. However, BCIs that use non-invasive external sensing, rather than brain implants have led to less precise control in the past. The current study uses novel sensing and machine learning techniques to access signals deep within the brain, achieving a high resolution of control, non-invasively over a robotic arm.
The current study utilises non-invasive neuroimaging and a novel continuous pursuit paradigm to overcome EEG signals leading to significantly improved EEG-based neural decoding, and facilitating real-time continuous robotic device control. The technology was tested in 68 able-bodied patients, using up to 10 sessions for each subject, including virtual device control and controlling of a robotic arm for continuous pursuit of a cursor on a computer screen.
Results show the non-invasive BCI successfully decoded neural signals,
allowing participants to control a robotic arm in real time, instructing it to
continuously and smoothly follow the movements of a cursor on a screen. Data findings show the new technique enhances BCI learning by nearly 60% for traditional centre-out tasks, and enhances continuous tracking of a computer cursor by over 500%. The team state their new framework addresses and improves BCI by increasing user engagement and training, as well as spatial resolution of non-invasive neural data through EEG source imaging.
The team surmise they have developed a successful, non-invasively
controlled, high resolution robotic device using only thoughts. For the future, the researcher state the technology is directly applicable to patients, and now plan to conduct clinical trials.
Source: Carnegie Mellon University