Research highlight

Neuroscience: Brains that work together

Scientific Reports

July 9, 2015

Two reports describing the use of brain-to-brain interfaces are published in Scientific Reports this week. In the first, a brain-machine interface that harnesses the brain power of multiple monkeys to complete a common task - controlling the movements of an avatar arm - is reported. The second discusses a brain-to-brain interface involving four adult rats capable of solving simple computational problems.

Brain-to-brain interfaces are networks formed by multiple animal brains that cooperate, allowing the animals to exchange sensory and motor information in real time.

In the first study, Miguel Nicolelis and colleagues recorded the activity of hundreds of neurons in the motor and sensory brain regions of four rhesus monkeys. Groups of two or three monkeys then sat in separate rooms with a computer screen that displayed an avatar arm. Their task was to move the arm towards a target using either a joystick (hand control) or by passively observing the movement of the arm, which was generated from the recordings of each monkey’s brain activity (brain control). The authors found that, with training, the monkeys using brain control increasingly coordinated their behaviour and their brain activity became more strongly correlated, leading to improved performance in the task.

In a separate study, the authors constructed a brain-to-brain interface called a Brainet, involving four rats, and investigated its ability to solve a set of computational problems. They implanted rats with multielectrode arrays in the primary somatosensory cortex to record neural electrical activity and transmit virtual tactile information to the brains of the other animals. Using different Brainet architecture designs, the rats were set a series of tasks including classifying two different stimuli and, on receiving temperature and barometric pressure information predicting an increased or decreased chance of rain. In the different tasks, the Brainet performance was equal to or better than an individual rat. Specifically, when the tasks required multiple computations, such as in memory storage or parallel processing, the Brainet performance was significantly improved when compared to individual rats.

doi: 10.1038/srep10767

Return to research highlights

PrivacyMark System