Distinguished decoding of the mind

The Getrud Reemtsma Foundation honours brain researchers Richard Andersen and Karl Friston with the International Prize for Translational Neuroscience

Richard Andersen (left) and Karl Friston, laureates of the International Prize f
Richard Andersen (left) and Karl Friston, laureates of the International Prize for Translational Neuroscience 2024. © private/ Kate Peters

A person steers a car through city traffic using only its thoughts, machines think and learn like humans - what sounds like something out of a science fiction novel could soon become reality thanks to current research. This year, the Getrud Reemtsma Foundation is honouring two researchers for their findings on the functioning of the brain. Richard Andersen from the California Institute of Technology is investigating how the brain processes sensory information and converts it into commands for movement. His research enables people suffering from paralysis to control their environment with their thoughts. Karl Friston from University College London is the founder of the  free energy principle. This states that the brain constantly makes predictions about the environment in order to avoid surprises. But it is not only the work of the brain that can be expressed in mathematical laws. The principle can also be applied to machine learning and could form the basis for an artificial intelligence that can think and learn like a human. The Translational Neuroscience Prize will be awarded on 20 June 2024 in Hamburg.

A man sits in front of a screen with green circles displayed on it. If a circle turns red, he moves the mouse pointer there quickly and precisely. Nothing unusual at first glance. However, the subject is completely motionless during the test. He has been paralysed from the neck down for years due to an accident and moves the mouse pointer with his thoughts alone. This feat was made possible by neuroscientist Richard Andersen from the California Institute of Technology and his team. Neuroscientists research how the 86 billion neurones in the brain work together to produce behaviour, including planning and executing movements.

Grasping an object requires a complex interplay of nerve cells. Visual cells in the eye are stimulated and transmit the image of the object to the brain as visual information. There, nerve cells are first activated in areas of the brain for vision and then in areas for movement. Finally, the information is transmitted from the brain to the spinal cord to activate the muscles in the arm and hand for gripping. But if the spinal cord is disrupted, as is the case with quadriplegics, they can still 'think' about movements but can no longer carry them out.

This is where the idea of a brain-machine interface comes in. Andersen and his colleagues use a chip measuring just four by four millimetres with 96 electrodes. It records when nerve cells in the brain are active and produce electrical signals. A computer decodes from the neural activity which movements are being planned and transmits a corresponding control signal that is used to move external aids.

Such neural prostheses could significantly improve the quality of life of quadriplegics. They make it possible to do much more than just control mouse pointers. Using the chip, participants can learn to operate a robotic arm that enables them to grasp independently. Even driving a car could become possible again for quadriplegics: In computer simulations, a test participant used his thoughts to steer an autonomous car through driving environments without any problems - and was as fast as non-disabled people in braking tests.

Free energy principle

Karl Friston from University College London has also been researching the processes and connections in the brain for years. He developed a unified brain theory that explains the complex processes in the brain using simple principles. According to his "free energy principle", all living beings strive for consistency and want to avoid surprises.

The brain doesn't like surprises either and is constantly making predictions about what might happen. This saves energy. If unexpected events occur, the predictions are adjusted and further improved. Surprises create awareness, which is stored as learning and experience.

However, the free energy principle can be used to explain more than just the thinking and learning processes in humans. The principle can also be applied to machine learning. Current artificial intelligence neural networks require a very large amount of data in order to learn the right patterns. Artificial agents that learn according to the principle of free energy can also operate in new and unknown environments. They act with the aim of avoiding surprises and learn from their prediction errors. On this basis, machines could one day be developed that think and learn in a similar way to humans.

The award winners

Richard A. Andersen studied biochemistry at the University of California, Davis, and he earned his doctorate in physiology from the University of California, San Francisco in 1979. He subsequently spent time as a postdoctoral research fellow at Johns Hopkins Medical School and as an assistant professor at the Salk Institute for Biological Studies in California. He became an associate professor in 1987 and then Professor of Neuroscience at the Massachusetts Institute of Technology (MIT) in 1990. Since 1993, he has been the James G. Bowsell Professor of Neuroscience at the California Institute of Technology (Caltech), where he has also heads the T&C Brain-Machine Interface Centre since 2017. He is a member of the National Academy of Sciences, the National Academy of Medicine, and the American Academy of Arts and Sciences.

Karl J. Friston studied Natural Sciences at Gonville and Caius College, Cambridge UK in 1980 and subsequently Medicine at King's College Medical School, London University. In 1988, he completed his specialist training in psychiatry at Oxford University and then went to Hammersmith Hospital, London for a research stay in the neuroimaging department. Between 1992 and 1994, he spent two years researching at the Neurosciences Institute La Jolla CA, USA. He has been a researcher at the Institute of Neurology, UK, since 1994 and has been Scientific Director of the Wellcome Trust Centre for Neuroimaging since 2001. Since 2022, he has also been the chief scientist at Verses, a company specialising in the development of new forms of artificial intelligence.