Brain prosthesis can recover lost memories

By PAIGE FRANK For The News-Letter

Science has created solutions for those who have lost arms, legs, hands and feet, but until now there has been no replacement for those who have lost memories.

This changed at the 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, held on Aug. 27, when the first-ever memory prosthesis design was announced. The device is the first attempt to create prosthetic technology to modify the function of the human brain.

The aim of the prosthesis is to aid those suffering from short-term memory loss, particularly Alzheimer’s patients. Many individuals suffering from Alzheimer’s seem to live in vivid memories of the past while unable to remember relatively simple facts such as what they had for dinner the previous night because the electronic signal the brain stores as a long-term memory is completely distinct from the original form the memory takes when it enters the brain.

The brain takes sensory input from a given event and transforms it into an electric signal representative of the event when memories are stored. That signal of the initial memory makes its way through regions of the hippocampus, the memory center of the brain.

Different regions of the hippocampus re-encode the signal as it is transported. The memory is represented by a completely new electronic signal designed for long-term storage upon reaching its final destination, and the memory cannot be safely stored until the signal has gone through multiple encodings. An individual with hippocampal damage (such as an Alzheimer’s patient) fails to form long-term memories because the hippocampus is unable to fully translate the memory into the proper form for long-term storage.

While it is called a prosthetic, the device is not designed to replace the damaged portion of a person’s hippocampus but rather to bypass the hippocampus all together, transforming short-term memory signals into signals that can be stored as long-term memories.

Theodore Berger and Dong Song headed the design of the prosthetic. They combined their efforts with data on neural activity collected by Sam Deadwyler and Robert Hampson of the Department of Physiology &Pharmacology of Wake Forest Baptist Medical Center.

The research began with a study of memory function in rats. Rats were trained to press two different levers in succession.

They were taught to press one lever and then, after a delay, to remember the initial lever and press the opposite lever. Two sets of minute electrodes were connected to the rats while they learned, obtaining neurological data of the activity in the left and right sides of the hippocampus.

Correct responses indicated that the rats had formed a short-term memory. When the rats were later retested, their short-term memory nerves were stimulated and, as expected, they performed better. The final step in the research involved injecting the animals with a nerve-blocking drug. Despite the influence of the drug, the rats remembered which lever to press when stimulated with the correct neural impulse pattern.

The next step in data collection examined volunteer patients who previously had electrodes implanted in their hippocampi to treat seizures. The electrodes in the patients enabled Hampson and Deadwyler to read the memory signals in the short-term memory region, as well as later in the long-term memory region. Using the collected neurological data, Hampson, Deadwyler, Berger and Song collaborated to construct an algorithm to mimic the function of the hippocampus. Hundreds of trials were conducted to determine the accuracy of the algorithm in predicting the long-term memory signal. While only nine different patients were used for the countless trials, the algorithm was accurate 90 percent of the time.

“Being able to predict neural signals with the USC model suggests that it can be used to design a device to support or replace the function of a damaged part of the brain,” Hampson said in a press release after revealing the results of the tests.

Designing that device is exactly what Berger then set out to do. The information acquired in the trials was used by Berger to model the process of memory signal transformation.w However determining the necessary signal to encode a long-term memory is only the first step. While Song and Berger were able to identify the signal that represents a short-term memory and determine how to create a long-term memory signal out of it, they had no way of knowing of what the memory they were transmitting actually consisted of.

“It’s like being able to translate from Spanish to French without being able to understand either language,” Berger said in a press release.

The challenge then became discovering a way to inject the translated memory signal back into a patient’s brain. To do this, the researchers designed a small array of electrodes to be implanted into the brain of a patient. In laboratory testing with animals the device performed well. Now it will be tested on human patients. If successful the device will create a signal that will serve as a long-term memory for patients with hippocampal damage.

“We’re not trying to understand the language,” Berger said. “Rather, on the basis of what we hear, can we translate something from Russian to Chinese without knowing either one?”