Four months later, without continued training, recall performance remained high.
Brain scans showed that memory training actually altered the brain functions of the trainees.
"After training we see massively increased performance on memory tests," says first author Dr Martin Dresler, assistant professor of cognitive neuroscience at Radboud University Medical Center in Nijmegen, the Netherlands.
"Not only can you induce a behavioural change, the training also induces similar brain connectivity patterns as those seen in memory athletes."
Dresler examined the brains of 23 world-class memory athletes and 23 people similar in age, health status, and intelligence but with typical memory skills.
Using structural MRI he measured differences in brain sizes.
He found there was no difference in brain anatomy between memory champions and normal people.
The differences they detected were in the connectivity patterns spread across the 2500 different connectors in the brain.
Memory athletes are not born with different brains.
"They, without a single exception, trained for months and years using mnemonic strategies to achieve these high levels of performance," Dresler says.
In this study, the strategy Dresler chose was memory of loci training.
Using this strategy, items on a list are associated with a remembered place, and users navigate that remembered place as they recall the list.
"Loci" is Latin for "places". If you want to remember a shopping list, for example, you imagine putting all the items in specific locations in a familiar place, such as your living room.
By using familiar places in the room to anchor the objects you will be able to recall them as you walk around the supermarket.
To explore the effects of training on the brain, Dresler and his colleagues recruited 51 individuals similarly matched to the memory athletes, but with typical memory skills and no previous memory training.
They were split into three groups: Two training groups and one group that did not train. The researchers scanned participants' brains before and after training.
The two training methods were short-term memory training and strategic memory training.
During short-term memory training, an individual practices remembering sequences, a bit like playing the game Concentration.
Strategic memory training provides trainees with a systematic way to remember lists.
Those who trained showed substantial improvement in their ability to recall lists of words.
Before training, individuals could recall on average between 26 and 30 words. Afterwards, those with strategic memory training could recall 35 more words on average.
Those who trained short-term memory could recall 11 more words. Those with no training recalled seven more words.
"Once you are familiar with these strategies and know how to apply them, you can keep your performance high without much further training," says Dresler.
After training, brain scans of those in the strategic training group had changed. They showed patterns that more closely resemble those of memory champions than scans taken before training.
To begin to understand how the connectivity patterns in the brains of memory athletes influence memory performance, Dresler and colleagues looked at the 25 connections that most differentiate memory athletes from others.
They found hubs of connectivity to two brain regions. One, the medial prefrontal cortex, is known to be active when individuals relate new knowledge to pre-existing knowledge.
The other, the right dorsal lateral prefrontal cortex, is known to be involved in efforts to learn strategically. "It makes sense that these connections would be affected," says Dresler.
"These are exactly the things we ask subjects to do when using method of loci for memorisation."
Dresler and his team are still analysing the brain scan data to learn more about the differences in brain connectivity patterns they found and how they affect memory.