LOS ANGELES — Applying a mild electrical current to a particular part of the brain could improve mathematical abilities in people who suffer impaired skills or in patients who have suffered strokes or other neurological problems, British researchers reported Thursday.

The weak current apparently allows neurons to fire more freely, stimulating the ability to learn, the researchers reported in the journal Current Biology. Reversing the flow of the current made it more difficult for neurons to fire, impairing learning ability.

The improvements in learning persisted for at least six months in the small study.

“”It’s perfectly conceivable that this type of stimulation can change people’s capacity to learn,”” said Dr. Edwin M. Robertson, a neurologist at Beth Israel Deaconess Medical Center in Boston who was not involved in the research. “”The stimulation techniques they are using, people think, can alter the capacity of circuits in the brain to change, and those changes underlie our capacity to learn.””

An estimated 6.5 percent of the population has a severe problem with even basic numerical understanding, a problem called developmental dyscalculia. As much as 20 percent have milder difficulties with math that present significant practical, educational and employment difficulties.

Neuropsychologist Roi Cohen Kadosh of the University of Oxford and his colleagues hope the technique they are developing can eventually be used to improve mathematical ability among such people, but they are still a long way from that goal.

The seat of mathematical capacity is the parietal lobes, which lie at the back of the head. Kadosh and his colleagues used a technique called transcranial direct current stimulation, in which a helmet applied to the head allows a very weak current to be passed through it. The current produces only a very slight tingling sensation, Kadosh said. Such stimulation has been studied for at least a decade for its potential to treat various neurological defects, including the after-effects of stroke.

The team worked with 15 volunteers ages 20 and 21, using a numerical variation of the well-known Stroop test. In the conventional Stroop test, the word “”red,”” for example, might be printed in green. An adult asked to identify the color of the word may hesitate longer than a child due to confusion between what he sees and what he reads.

The mathematical version assigns numerical values to shapes previously unseen by the participants. The researchers might then ask them which of two shapes has a large value, confusing them by making the picture of the shape with the larger value smaller than the picture of the shape with the smaller value.

Passing a current through the subjects’ brains in one direction while they were learning the task improved their ability to mentally manipulate the shapes, while passing the current in the opposite direction impaired it. Six months later, the difference was still apparent.

Left unanswered is whether this type of stimulation can help people with dyscalculia, Robertson said. Testing it in children with the disorder will be especially tricky.

“”We always have concerns about using children in any experimental situation,”” he said. “”They are a vulnerable group, it is difficult for them to give consent and their brain is still developing. We have to move gingerly toward that aim.””