Researchers implanted a device into mice which contained an LED light and a chamber full of cells whose gene expression could be controlled by volunteers' brain activity.
Brainwave activity from volunteers wearing wireless headsets switched the light on. This led to expression of a light-responsive gene in the genetically-engineered cells, and thus production of a protein, which was released into the bloodstream of the mice.
They hope their experiments could represent the first step towards a system allowing people who are fully conscious but have difficulty communicating - such as in locked-in syndrome - to self-administer their medication.
'Being able to control gene expression via the power of thought is a dream that we've been chasing for over a decade,' said Professor Martin Fussenegger, from ETH Zurich in Basel, Switzerland, and lead author on the study.
In the study, the researchers were able to modulate the amount of protein released by changing the headset-wearing volunteers' state of mind – for instance making them concentrate, or relax. Ultimately, the volunteers learnt to directly control the LED light by will alone (a control technique known as bio-feedback).
'In all of these different mental states, we saw very specific brain activities and these were translated via the LED to very specific illumination of the designer cells. In response those [genes] produced proteins that were then circulating in the animal,' Professor Fussenegger told BBC News.
In addition to a potential application in locked-in syndrome, Professor Fussenegger says the technique could be used to detect aberrant brainwave patterns in neurological diseases, triggering a rapid treatment response. He told New Scientist that he hopes to start clinical trials within ten years.
But Professor Geraint Rees of the Institute of Cognitive Neuroscience at University College London, who was not involved in the study, told BBC News that interpreting brainwaves is very difficult, which might limit the technology's usefulness. 'We don't yet have a complete understanding of how to translate what we can record on the surface of somebody's scalp into a pattern of thought,' he said.
The study was published in Nature Communications.