Exploring how the brain predicts the future

Experience changes the way we behave. This is also true for the neurons in our brain. Scientific evidence shows that our brain uses past experience to infer what is to be expected for the immediate future. Now, ESI scientist Martin Vinck has been awarded with an ERC starting grant to research how this is implemented by brain cells.

3 Sep 2019


Crystal balls, tarot cards, tea leafs – fortune telling methods that are most evidently ineffective. If there’s one thing that might be able to predict the future, it is the brain. “The ability to predict is critical, because actions do not take place in the immediate present, but in the future.” says Martin Vinck, principal investigator at the Ernst Strüngmann Institute in Frankfurt am Main. “For example, a tennis player needs to prepare a racket swing while the ball is still in the other side of the tennis court, and thus needs to predict the future trajectory of the ball.” The ability to infer future situations based on prior experience very likely is an important key to how we perceive the world. To better understand how this strategy is implemented in the brain Martin Vinck now receives one of the prestigious ERC starting grants offered by the European Commission to support excellent research by young scientists.

Looking forward to first results: Martin Vinck is excited about the opportunities the ERC grant is offering. (picture credit: Martin Vinck)

Looking forward to first results: Martin Vinck is excited about the opportunities the ERC grant is offering. (picture credit: Martin Vinck)

The world as perceived by the brain

In a traditional view the brain works in a highly hierarchical manner. Sensory cells are specialized to detect certain properties of a stimulus, let’s say the image of a chair. Some cells will register its outline, others its colour. Higher order cells in the next level of processing bring these information together and pass it on again until eventually a coherent image is constructed, attached with a name tag and meaning – based on which a behavioural decision can follow, such as sitting down. While this systematic approach explains well how perceptions of objects are constructed, there are reasons to suspect this might not be the default working mode of the brain: We recognize a chair as a chair also when its defining characteristics are skewed, such as standing upside down or when partly hidden by another object. Moreover computing an image bottom up from its smallest details every time you encounter it takes a lot of time. Time that can be precious in an environment where survival depends on quick actions.

Perceiving our environment gets a lot quicker, more efficient, and flexible when the brain makes use of the experience it has about the world. Much like listening to the first tune of a song and already knowing what will follow, the brain has the ability to infer from a minimum amount of sensory evidence what it has to expect for the larger picture. This way standard templates can be filled in without effort and time: Brown object in known location – chair in front of office desk – sit down. Capacity then is freed up to process relevant information, namely aspects of the world that are not as expected. Brown object in known location making roaring noises – grizzly in front of office desk – run away.

Unraveling the secrets of neural communication

Over the past twenty years scientists have found more and more evidence for the predictive abilities of the brain. However very little is known about how this strategy is implemented by the neuronal cells. In his ERC funded project Martin Vinck hopes to change exactly that: “In any given situation, some sensory inputs could be predicted from prior experience, while others could be unpredicted. We are testing exactly how the responses of neurons and their coordinated activity is modulated by this predictability.” He expects the timing of neuronal responses and particularly the rhythm that is produced by the simultaneous activity of multiple neurons will be critical to separate predictable from unpredictable inputs. To test these ideas he and his co-workers will use computer algorithms to detect response rules of neurons and predict patterns of brain activity.

The idea for the project has been on Martin Vinck’s mind for quite some time. The ERC grant now puts him in the position to finally pursue his idea. Needless to say he is thrilled: “The project is challenging but I am very much looking forward. I can’t wait for the first results to come in.” What these results will look like no brain can predict. But irrespective of its exact findings, the project surely will increase our understanding of the way the brain works.

Prospective PhD Students and Postdocs who are interested in working with Martin Vinck on this project are welcome to get in touch with him via email. Further information about him and his projects can be found on his homepage.