Researchers have developed the first memory prosthetic device—a neural implant that, in rats, restored lost brain function and improved short-term memory retention. While human testing is still a distant goal, the implant provides evidence that the brain’s complex neural code can be interpreted and reproduced to enhance cognitive function.
The device, which consists of a tiny chip and a set of 32 electrodes, marries math and neuroscience. At its heart is an algorithm that deciphers and replicates the neural code that one layer of the brain sends to another. The function restored by the implant is limited—rats were able to remember which of two levers they had pressed. But its creators believe that a device on the same principle could one day be used to improve recall in people suffering from stroke, dementia, or other brain damage.
Wake Forest University neurophysiologist Samuel Deadwyler first trained the rats to press two different levers in succession. The animals learned to press one lever as it was presented to them and then, after a delay, remember which they’d pressed and choose the other one the second time around. While the rats performed the task, two sets of minute electrodes recorded the activity of individual neurons on the right and left sides of the hippocampus, an area of the brain that consolidates short-term memory by processing information as it passes through multiple layers. A set of 16 electrodes—eight on the right, eight on the left—monitored signals being sent from neurons in an area of the hippocampus called the CA3 layer, and another 16 monitored the processed signals received by neurons in the CA1 layer.
Together with Theodore Berger, a biomedical engineer and neuroscientist at the University of Southern California, Deadwyler characterized the pattern of neural activity associated with a correct response—the pattern indicating the formation of a solid short-term memory. The researchers stimulated the nerves in the same pattern and retested the rats. This time, the animals made fewer mistakes and could remember which lever to press even after longer delays. When the researchers took it a step further, preventing memory formation with a nerve-blocking drug, they found that the rats could still “remember” which lever to press if they were stimulated with the neural impulse pattern.
“It’s an exciting demonstration of the capabilities that we have now, not of only reading neuronal activity of the brain but also manipulating it,” says Charles Wilson, a neuroscientist and emeritus professor at the University of California, Los Angeles, who was not involved in the research. “Hopefully, this could be clinically useful in the future.”
Part of the challenge in creating the prosthesis was to develop a device that would ultimately be able to assist in the recall of many types of memories. That required learning to replicate the activities of the hippocampus. Rather than storing specific memories, the hippocampus passes them along to the brain’s long-term memory, translating them into a form that the long-term memory is able to store. Similarly, the algorithm does not store specific examples—how to brush your teeth, how to find your way home—but instead creates a set of rules much like the ones a voice-recognition program might use to translate one language into another. “We’re not trying to understand the language,” says Berger. “Rather, on the basis of what we hear, can we translate something from Russian to Chinese without knowing either one?”
Berger and Deadwyler are now working to increase the number of neurons they can monitor and to move their research into nonhuman primates—next steps on the long journey toward developing a human implant. “We already have the technology and capability to record and stimulate a single neuron in humans; the ingredients are already there,” Wilson says. “And the fact that it could be done in animals suggests to me that a similar thing could be done in humans.”
A quick guide to the most important AI law you’ve never heard of
The European Union is planning new legislation aimed at curbing the worst harms associated with artificial intelligence.
It will soon be easy for self-driving cars to hide in plain sight. We shouldn’t let them.
If they ever hit our roads for real, other drivers need to know exactly what they are.
Crypto is weathering a bitter storm. Some still hold on for dear life.
When a cryptocurrency’s value is theoretical, what happens if people quit believing?
Artificial intelligence is creating a new colonial world order
An MIT Technology Review series investigates how AI is enriching a powerful few by dispossessing communities that have been dispossessed before.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.