A "brain chip" could be used to replace the "memory centre" in patients affected by strokes, epilepsy or Alzheimer's disease, it has been claimed.

US scientists say a silicon chip could be used to replace the hippocampus, where the storage of memories is coordinated.

They are due to start testing the device on rats brains shortly.

If that goes well, the Californian researchers will test the artificial hippocampus in live rats within six months and then monkeys trained to carry out memory tasks before progressing to human trials once the chip has been proved to be safe.

The aim is that the silicon chip - the first brain prosthesis - will be able to replace damaged brain tissue. Current devices, such as cochlear implants, only stimulate brain activity.

The researchers are aware that their work could provoke controversy because the brain affects mood, awareness and consciousness, as well as memory - areas directly linked to a person's identity.

But Theodore Berger, from the University of Southern California in Los Angeles, who has been leading the research, said this was a test case.

He said: "If you can't do it with the hippocampus you can't do it with anything."

The hippocampus is an area at the base of the brain in humans, close to the junction with the spinal cord.

It is believed it "encodes" experiences so they can be stored as long-term memories in another part of the brain.

Dr Berger told New Scientist magazine: "If you lose your hippocampus you only lose the ability to store new memories.

"That offers a relatively simple and safe way to test the device. If someone with the prosthesis regains the ability to store new memories, then it's safe to assume that it works."

The researchers have spent 10 years developing the artificial hippocampus.

Scientists do not know exactly how the hippocampus works.

So the Californian team simply copied its behaviour.

Slices of rat hippocampus were stimulated with electrical signals millions of times, until scientists could be sure which input produced a corresponding output.

Putting the information from each slide together, the researchers were able to devise a mathematical model of a whole hippocampus.

The model was then programmed on to a chip.

They suggest the chip would sit on a patient's skull, rather than inside the brain.

It would communicate with the brain via two arrays of electrodes, placed on either side of the damaged attitude.

One would record the electrical activity coming from the rest of the brain, while the other would send out the necessary instructions back to the brain.

The researchers say that, because the hippocampus can be seen as a series of similar circuits that work in parallel, it should be possible to bypass the damaged area.

They are shortly going to carry out the tests on the slices of rat brains kept alive in cerebrospinal fluid.

Dr Berger said: "It's a very important step because it's the first time we have put all the pieces together."

Once those trials are complete, the researchers will begin tests on monkeys where they stop part of the hippocampus working and by-pass it with the chip.

Sam Deadwyler, of Wake Forest University in Winston-Salem, North Carolina, who will conduct the animal trials commented: "The real proof will be if the animal's behaviour changes or is maintained."

Bernard Williams, a philosopher at Oxford University who is an expert in personal identity, said people may find the technology hard to accept at first.

But he added: "Initially, people thought heart transplants were an abomination because they assumed that having the heart you were born with was an important part of who you are."

Making a difference

Dimitri Knullmann, professor of neurology at the Institute of Neurology, told BBC News Online the chip was "miles away" from being used in human brains.

He said the researchers appeared to have concentrated on the electrical signals going in to and out of the hippocampus, and not the other complex activity seen within the area.

Professor Knullmann added: "I think they would run into major difficulties in testing this in any sensible way because taking out the existing hippocampus and wiring this device in somehow would cause damage.

"Proving conclusively the chip, as opposed to just leaving a hole, makes a difference is going to be difficult."

The research is to be presented to a neural engineering conference in Capri, Italy, next week.