11-21-2015, 06:39 AM
http://journals.plos.org/plosone/article...ne.0131328
There have been brain-computer interfaces before, where researchers carefully located sites in the brain to place electrodes, corresponding to motor centers where very similar tasks were being controlled already. And the brain eventually learns to use these interfaces to accomplish closely related tasks - but they don't tend to be stable in the long term because brain plasticity. Unless the task is truly part of the routine on such a frequent basis that the artificial interface rate the support of long-term learning and adaptation and myelination of the neurons leading to it - the selected sites cease to be the *exact* sites that the brain uses for the related tasks.
Now there is an Interesting new feature, enabled by better sensing: Using adaptive algorithms to figure out what parts of the brain ought to be paid attention to when using an artificially controlled system to accomplish a task.
It's a lot more general, and can in principle be applied modally enabling a trained brain to learn *NEW* brain-controlled hardware-assisted tasks without further hardware.
There have been brain-computer interfaces before, where researchers carefully located sites in the brain to place electrodes, corresponding to motor centers where very similar tasks were being controlled already. And the brain eventually learns to use these interfaces to accomplish closely related tasks - but they don't tend to be stable in the long term because brain plasticity. Unless the task is truly part of the routine on such a frequent basis that the artificial interface rate the support of long-term learning and adaptation and myelination of the neurons leading to it - the selected sites cease to be the *exact* sites that the brain uses for the related tasks.
Now there is an Interesting new feature, enabled by better sensing: Using adaptive algorithms to figure out what parts of the brain ought to be paid attention to when using an artificially controlled system to accomplish a task.
It's a lot more general, and can in principle be applied modally enabling a trained brain to learn *NEW* brain-controlled hardware-assisted tasks without further hardware.