A pioneering neuroscientist who succeeded in enhancing memory in rats using brain chip implants is now ready to start trialling the technology on humans, spinning off his research into a startup to make the implants into a commercial product.

Dr Theodore Berger, a professor of biomedical engineering at the University of Southern California (USC) in the US and director of USC's Center of Neural Engineering, has spent his career trying to figure out how neurons form memories, in order to help people suffering from Alzheimer's disease, dementia, stroke or brain injuries restore the ability to create long-term memories.

To that end, over the last 20 years he has been developing brain prostheses – computer chips that mimic the electrical signal processing that occurs in parts of the hippocampus, which is a major component of the human brain and one of the first regions in the brain to suffer damage in Alzheimer's.

In order to convert a short-term memory into a long-term one, the human brain fires a set of electrical signals in a code that is unique to each person. When brain cells are damaged, they are not able to do this, so Berger pioneered software that gets the brain chip to fire off a specific set of electrical signals matching the existing code pattern in the brain.

Over the last few years, Berger and his colleagues have published several studies showing how they have been able to enhance memory in both live rats and monkeys that had the chip implanted into their brains.

The Defense Advanced Research Projects Agency (Darpa), which funded some of the studies, has a keen interest in brain chips as it wants to create computer chips to help wounded soldiers heal, as well as making a brain-computer interface possible.

Berger has spun off the research into a new startup called Kernel, which is backed by technology entrepreneur Bryan Johnson, who sold his payments firm Braintree to PayPal for $800m (£615m) in 2013.

Johnson has a particular interest in artificial intelligence and, according to the Washington Post, he hired a team of neuroscientists to analyse all current research into evolving the human brain and pick the best scientists that might be interested in setting up a company working on brain technologies.

In the end the team of neuroscientists chose Berger, who received a cold call from Johnson, and 10 months later the Kernel team are working on prototypes of the brain implant device and are starting to conduct tests with epilepsy patients in hospitals.

"We're testing it in humans now, and getting good initial results," Berger told IEEE Spectrum. "We're going to go forward with the goal of commercialising this prosthesis."

The most immediate problem with creating viable a brain prosthesis will be making Berger's technology portable, and at the moment the epilepsy patients in the trial have to have temporary electrodes placed in their brains.

The other issue will be being able to make the algorithms good enough to always correctly predict what the unique code is in each human brain so it can mimic the electrical signals and ensure that memory recall is always consistently successful.