Among the hardest challenges for certain stroke patients may well be the inability to find and speak words even if they are aware of what they want to say.
Speech therapy is laborious and can take months. New research is seeking to cut that time significantly, with the help of non-invasive brain stimulation.
"Non-invasive brain stimulation can allow painless, inexpensive, and apparently safe method for cognitive improvement with with potential long term efficacy," said Roi Cohen Kadosh of the University of Oxford.
A focus of many of these studies is tDCS - transcranial direct current stimulation.
In tDCS, researchers apply weak electrical currents to the head via electrodes for a short period of time, for example 20 minutes. The currents pass through the skull and alter spontaneous neural activity.
Some types of stimulation excite the neurons, while others suppress them. Subjects usually feel only a slight tingling for less than 30 seconds.
The effects of tDCS can last for up to 12 months, Cohen Kadosh said, "most likely due to molecular and cellular changes that are important mechanisms implementing learning and memory."
For Jenny Crinion of University College London, who is both a neuroscientist and clinical speech and language therapist, the interest in tDCS sprang from a desire to help stroke patients through their long recovery.
While speech therapy works well at improving speech following aphasic stroke, it can be frustratingly slow.
She hopes to pair brain-stimulation interventions with proven language-rehabilitation methods, Crinion said, "such that the same maximum recovery is ultimately achieved as with therapy alone but with fewer hours of rehab."
Crinion's current work focuses on understanding how tDCS affects the areas of the brain involved in speech production.
She paired an fMRI picture-naming study with a 6-week-long tDCS and word-finding treatment study to see if brain stimulation could improve stroke patients' speech both immediately after treatment and three months later.
In the picture-naming task, people were presented with pictures of simple, everyday words such as car and asked to name them as quickly and accurately as possible.
The results support other studies that tDCS can speed up word finding in both healthy older people and stroke patients, and are helping to identify which parts of the brain should be stimulated.
At the same time, she cautioned, one type of treatment may not fit all patients, and further work will clarify whether some patients may also benefit from treatments targeted at the brain hemisphere not affected by stroke.
In a different set of experiments that look at the effects of brain stimulation on memory, Paulo Sergio Boggio of Mackenzie Presbyterian University in Sao Paulo, Brazil, used tDCS to try to enhance the memory of Parkinson's and Alzheimer's disease patients.
In his study of Alzheimer's patients, Boggio tested how many sessions of tDCS would lead to sustained improvements in memory and visual recognition.
His team used five consecutive sessions of tDCS to excite two different areas of the brain involved in motor planning, organization, and regulation.
Visual recognition increased by as much as 18 percent in the Alzheimer's patients, and the effects lasted a month. In a similar study with Parkinson's disease patients, tDCS improved memory by 20 percent.
Cohen Kadosh of the University of Oxford studied a very different application of tDCS - how to improve how people learn about numbers.
Citing a recent study that found that approximately 20 percent of British adults have numeracy skills below the minimum requirement for being fully functional in the modern economy, Cohen Kadosh explains that there is currently no solution for low numerical abilities, aside from behavioural training.
"I believe that this is an important problem with many implications for society," he said.
His studies have found that it is possible to enhance numerical abilities using tDCS applied to the part of the brain called the posterior parietal cortex. The observed improvements lasted up to 6 months after tDCS and were specific to the trained material.
Cohen Kadosh has also tested the effects of tDCS on people with low numerical abilities due to congenital factors - dyscalculia, the equivalent to dyslexia with numbers that affects about 5 percent of the population.
For those individuals, tDCS was only effective if it targeted different regions of the brain than those in people without dyscalculia.
"This suggests that people with dyscalculia recruit different brain areas for numerical processing, probably due to brain reorganization," he added.