The race for a human memory-aid prosthesis

The time target is four years. The goal is to develop the first memory-enhancing prosthesis, to be implanted in the brains of people with impaired memory function caused by cerebral trauma.

November 2014

IMA Lab The race for a human memory-aid prosthesis

Since the year 2000, over 270,000 U.S. military personnel have suffered brain damage affecting their memories, and no satisfactory pharmaceutical treatment has yet been found.

In response, the U.S. Defense Advanced Research Projects Agency (DARPA) has commissioned the University of Pennsylvania’s Computational Memory Lab to carry out further studies on the neuroscience of memory, with the aim of developing a physical prosthetic memory-aid to implant in the human brain: a challenging goal, made all the more ambitious by its timetable, which establishes a completion deadline in just four years.

The research team will be directed by Michael Kahana. It will commence by searching for the biological markers connected with memory formation and retrieval. This initial research will be conducted using hospitalized epilepsy patients as test subjects, since they have already had brain electrodes implanted (to allow doctors to study their epileptic seizures). Kahana and his team will record the electrical activity in these patients’ brains while they carry out various memory tests.

“Human memory works like a search engine,” Kahana says. “In the initial memory encoding phase, each event has to be tagged. In the memory retrieval phase, you need to be able to use those tags to search effectively.” He aims to identify the electric signals associated with each of these two operations.
Once these have been found and understood, researchers will try to amplify them using sophisticated neural stimulation devices capable both of recording neural activity and of stimulating the brain to recover it effectively.

Neural specialists worldwide have long contemplated and desired such a “closed-loop” device: a prosthesis that could monitor and use real-time signals from the brain to define stimulation parameters for data storage and retrieval… but the challenges are immense, not least the major chip engineering challenges. In particular, the “incoming” recorded information must be interpreted with staggering speed for it to be translated instantly into a stimulation command. This will mean completing in 10 milliseconds operations that would take a normal personal computer several hours to carry out.

A second team, led by Itzhak Fried, director of the University of California’s Cognitive Neurophysiology Laboratory, will focus on a part of the brain called the entorhinal cortex, a kind of gateway to the hippocampus, the primary brain region associated with memory formation and storage. “Our approach to the RAM program is focused on homing in on this circuit, which is really the golden circuit of memory,” states Fried who, in an experiment in 2012, demonstrated that stimulating the entorhinal regions of patients while they were learning memory tasks, notably improved their performance.

Both teams will liaise with the Livermore Center for Bioengineering, whose researchers are leveraging semiconductor manufacturing techniques to make tiny implantable systems.
The memory prosthesis will consist of two flexible cylinders about 1 millimeter in diameter, each studded with up to 64 hair-thin electrodes, which will be capable of both recording the activity of individual neurons and stimulating them.

The various research units involved in this project are all optimistic about the prospect of achieving the goals set by DARPA in only four years… but at the same time, no one underestimates the immense difficulty of the challenge facing them. The scientific world will be watching.