Neuroscientists from the Massachusetts Institute of Technology have designed a new computational model that they claim explains how the brain can learn very similar tasks without mixing them up, while maintaining the balance between plasticity and stability.
The brain is made up of billions of neurons, all of which connect to others. According to the researchers, on average each neuron connects to about 10,000 others.
The researchers, who write in the journal Proceedings of the National Academy of Sciences, say this connectivity is key.
Neurons are constantly changing their connections, forming new patterns or strengthening existing ones. This plasticity allows us to learn new tasks, such as perfecting a golf swing or improving a tennis stroke. And even though not all the connections will be relevant to the task, they allow the brain to explore new ways of achieving the goal.
As the brain learns what is required for a new motor skill, neurons form "circuits" to produce the goal - for example, moving the body to swing a golf club. And as it is unlikely that the desired outcome will be achieved the first time, feedback from the neurons enables the brain to seek alternatives.
Neurons can become specialized for specific tasks, the researchers claim.
Robert Ajemian, a research scientist at the McGovern Institute for Brain Research at the Massachusetts Institute of Technology (MIT) and lead author of the paper, explains:
"Your brain is always trying to find the configurations that balance everything so you can do two tasks, or three tasks, or however many you're learning. There are many ways to solve a task, and you're exploring all the different ways."
This theory works well if you are just learning one skill, but it presents complications if you are trying to master more than one thing at once. As the same network controls related motor skills, any changes to the existing "circuits" may interfere with previously learned skills.
The researchers claim connectivity is advantageous, as it allows the brain to test different solutions to achieve combinations of tasks.
The constantly changing connections of the neurons - called hyperplasticity - is balanced by another inherent trait: they have a very low signal-to-noise ratio, which means they do not filter useful from useless information coming from their neighbors.
The researchers factored the noise into their computer model, as they believe it is critical to the brain's learning ability.
"Most people don't want to deal with noise because it's a nuisance. We set out to try to determine if noise can be used in a beneficial way, and we found that it allows the brain to explore many solutions, but it can only be utilized if the network is hyperplastic."
Without the noise, the scientists say, old memories could be easily overwritten and the learning lost. On the other hand, without the plasticity, new connections and learning would not be possible, as the tiny changes would be drowned out by the background noise.
The researchers claim that measuring the growth and formation of connections of dendrites - the tiny "extensions" that neurons use for communication - provides anatomical evidence that neurons demonstrate plasticity even when learning is not taking place.
Ajemian points out that the constantly changing connections explain why some skills are lost if they are not practiced - especially if they involve skills that overlap with other learnings.
That is why, Ajemian says, athletes warm up before a game - not just to wake up muscles but to "flush" the neural pathways to the brain's motor cortex and enable signals for learned strokes free passage.
However, some skills, such as riding a bike, are not so easily forgotten, as the neural connections do not overlap as much.
"Once you've learned something, if it doesn't overlap or intersect with other skills, you will forget it but so slowly that it's essentially permanent," Ajemian says.
The researchers are now exploring whether the model can explain how the brain forms memories of events, not just motor skills.