Research at ELSC

Minds in bodies: Sensation, perception and awareness in health and disease

Minds and bodies seem to live in very different realms, yet are (almost) inseparable – this is the origin of the mind-body problem. ELSC scientists conduct a concerted effort to address the connection between minds and bodies by studying both the nature of perception by minds on the one hand, and their underlying biological foundations on the other.

ELSC researchers are addressing questions such as: How do we conjure up and consciously perceive a coherent model of the world around us, based on indirect and complex information that reaches our sensory organs in the form of photons, chemicals, pressure waves, and other sensory stimuli?  Why does brain disease, from chronic pain and stroke to degenerative brain diseases and schizophrenia, alter the fluency of this interaction with the world? 

These investigations into Sensation and Perception may make it possible to greatly improve the quality of life of millions of people. Doing so begins with the scientific research as done at ELSC, involving the study of sensory processing at the levels of single neurons, neuronal circuits and whole-brain areas, as well as the complex interactions between different areas of the brain that ultimately generate conscious perception.

From decision to action: How brains generate and control movements

Our remarkable ability to move in space and interact with the environment requires the brain to process information at astonishing speed.  We can appreciate the complexity of our motor system by comparing it to artificial devices. While today’s computers outperform humans in tasks previously considered as the pinnacle of human intelligence, such as playing chess, the human motor capability is superior to any existing robotic system. The complexity of the motor system and the computations it carries out are further revealed when the system is impaired. Disorders of motor control triggered by illness (e.g., Parkinson’s disease) or injury (e.g., spinal cord injury) are devastating, affecting almost all forms of behavior, and the current therapeutic techniques are limited.

Voluntary movements are formed by an orchestrated activation of many muscles to achieve a goal. However, muscle activation is only the final outcome of multiple processes that include a decision to move, an evaluation of alternative outcomes, the planning of a strategy, and a final execution. These processes are implemented by a dynamical flow of information through multiple feedback loops implemented by interactions within and between many brain centers. This situation imposes a great challenge for studying the transformation of decisions into actions. Addressing this challenge requires a collaborative, integrative, and multidisciplinary approach.

ELSC’s motor control research program aims to crack the motor code. This research program spans different motor systems from the simple zebrafish larva to the complex human system, providing a unique opportunity to transfer and integrate knowledge aiming to extract general rules governing the motor system. Scientists at ELSC seek not only to improve current therapeutic techniques, such as deep brain stimulation, but also to work toward the development of novel approaches for treating motor disorders through the identification of new targets for monitoring and intervention, including the development of better brain-machine interfaces.

Artificial and biological intelligence: Moving artificial intelligence to the next level

Recent technological advances have revolutionized the field of artificial intelligence (AI). Big databases and enhanced computing capabilities have allowed for the construction of AI systems that for the first time can solve real-life problems in vision, speech and language, in some cases paralleling the capabilities of humans. Many of these advances are based on brain-inspired learning algorithms called deep learning (DL). Nevertheless, the current AI is still a far cry from human intelligence in flexibility, depth, breadth and speed of learning. These developments pose new and exciting opportunities and challenges to neuroscience. Addressing them will shape brain research in the decades to come.

While DL has been successful in real-world applications, it is a challenging question whether DL is a viable model of brain function. For example, current robotic systems are still extremely limited in their ability to navigate and act freely in complex environments, while image and speech recognition systems based on DL are open to attack by adversarial examples that would hardly fool a human observer. Conversely, it is challenging to inform AI about novel principles of learning and intelligence that emerge from neuroscience. The increasing exposure of individuals and the public to AI in everyday devices brings into sharp focus fundamental questions about the nature of intelligence, consciousness, freedom and autonomy.

ELSC researchers are developing a theoretical understanding of the successes and limitations of DL, testing the extent to which DL is a viable model of brain function, and moving AI to the next level by distilling our understanding of brain function into novel principles of learning, knowledge representations, computation and intelligence.

The diseased brain: Characterization and repair

Brain disorders contribute increasingly to the total prevalence of disease in the world. Neurological disorders are the third most common cause of disability and death in Europe (and probably in the rest of the developed world). Although the burden of brain disorders increases with age, they do occur at all ages: Neurodevelopmental atypicalities such as autism and dyslexia manifest in childhood; schizophrenia and other psychiatric diseases tend to occur in young adults, while addictions can occur at all ages; and stroke, Alzheimer’s disease, and Parkinson’s disease are disabling neurological disorders that tend to hit older populations. Disorders linked to brain function represent a high societal and economic burden and lead to an increasing demand for new and better treatments.

At ELSC, our researchers are tackling these and similar issues in order to characterize the typical and atypical brain, to understand brain disorders, and to develop treatment and repair strategies. Cognitive neuroscientists collaborate with theoretical neuroscientists to develop sophisticated computational techniques to characterize the atypical cognitive characteristics of dyslexia and autism, while molecular biologists characterize the molecular signatures of addiction. ELSC researchers have reproduced brain pathologies in organelles that develop from pluripotent stem cells, and have conducted pioneering characterization of Alzheimer’s disease pathologies beyond neuronal degeneration. By further development and improvement of deep brain stimulation techniques to treat Parkinson’s and other movement disorders, ELSC researchers have improved the lives of countless patients. Using the strong localization of some language functions, linguists at ELSC are providing tools for following the time course of a stroke.

These exciting results highlight the interdisciplinary tools and the collaborative ethos that characterize ELSC, where biologists, cognitive neuroscientists and theoretical neuroscientists join forces to push the frontiers of brain treatment and repair.

“Working memory”