Action Lab
Research

What is our research interest?

The selection of actions and appropriate responses is a fundamental problem every organism is faced with. Even though there are hundreds of response alternatives that are accessible to us every moment, we are almost effortlessly able to choose between different alternatives to organize our behavior. We are faced with this problem in many areas of daily life.  For example, when learning to drive a car it is challenging to coordinate steering, braking, switching gears, and tuning the radio at almost the same time. Despite we are faced with such problems in many situations, the question of how we choose between different response alternatives is not satisfactorily answered. Moreover, the question of how we are able to control our actions in some situations, but not in others is also not answered.  

The goal of our research is to combine cognitive science approaches with (applied) neuroscientific research through the use of brain stimulation methods,  computational neuroscience and artificial intelligence to unravel how humans control their actions: From a basic neuroscientific perspective, we aim to unravel the neurobiological and neurophysiological mechanisms underlying  action control. Our goal from an applied point of view is to establish a multifactorial framework to explain the cognitive patho-neurophysiology of fronto-striatal and fronto-parietal networks responsible for response selection in our brain. Our research is organized by three inter-related branches (i.e. basic science, neural technology and clinical science) as shown in the figure below.

What neuroscience methods do we use?

The research concept integrates different, yet interlinked approaches in healthy people and neuropsychiatric disorders. Major  research methodologies include

• • EEG

  • MRI

  • EEG/TMS

  • tDCS/tVNS/TMS

  • computational modeling

  • artificial intelligence (CNNs, RNNs, GANs etc.)

Besides basic neuroscientific research, we are also interested in how brain mechanisms mediate action control under ‘pathological’ conditions. To understand the neuronal mechanism at the deepest possible level, results from the above-mentioned studies in healthy and diseased people are simulated in neural networks using "computational neuroscience" techniques, including artificial intelligence methods. Combined with brain stimulation techniques the goal is to gain causal, mechanistic insights into the neurophysiological underpinnings of human action control and disturbances thereof. 

Clinical conditions we are interested in

Clinical conditions focused refer to

• • Tourette Syndrome

  • ADHD

  • Autism Spectrum disorder

  • Obsessive-Compulsive disorder

  • neurodegenerative disorders

  • various neuropediatric diseases

  • drug abuse and addiction

Concerning these diseases, we are interested in how multi-tasking and action selection processes are changed as the disease progresses. As a part of these disease-related approaches, we also investigate the effectiveness of treatments. Besides the effects of 'classical' drug-based interventions, we investigate the effects of 'deep brain stimulation (DBS)' and non-invasive brain stimulation methods. Results of these approaches provide insights that are of relevance from an applied perspective since these studies contribute to the knowledge necessary to optimize treatments for these patient groups. At the same time, these approaches provide deep insights into the mechanisms of action control. 

Translation into healthcare

We are part of SaxoChiLD, a regional network comprising the Technische Universität Dresden, the Universität Leipzig, the Robert Koch Institute, the Max Planck Institute for Evolutionary Anthropology, The Helmholz Center for Environmental Research and the Helmholz Institute for Metabolic, Obesity and Vascular Research in Germany. The overarching approach of SaxoChiLD is comprehensive epidemiological surveillance of child health and development within a changing environmental context, with special consideration of the interaction between psyche and soma.

In this context, the chair deals with the properties of neural networks for the evaluation of complex high-dimensional data (e.g. EEG data). He combines fundamental expertise in the functions of neural networks and their technical implementation for the evaluation of data. As part of BMBF projects, the chair is also developing methods for the EEG classification of diseases based on AI-supported procedures for analyzing medical reports.