Updated information will come soon ...

 

Research on Functional Imaging Biomarkers

Complex disorders of the human central nervous system like schizophrenia and autism, for which biological validity of diagnosis is not yet established, constitute a big challenge for drug development. Given the lack of biological validity, animal models only recapitulate some biological or phenomenological aspects of these diseases. Therefore, the efficacy of a drug is often only properly assessed at late stage clinical trials. Functional imaging biomarkers offer the possibility to close the gap by providing a read-out of brain activity for relevant cognitive functions in humans, and its modulation by drugs, throughout the whole cycle of clinical trials. Functional MR imaging (fMRI) constitutes one of the most commonly used methods for non-invasive functional imaging of the human brain with the highest spatial resolution. New fMRI data analysis approaches provide increasingly sensitive measures for assessing brain function.

 

Motor Control Research

Human arm movements can be incredible fast and accurate, despite their versatility and the complex neural processing underlying these actions. How does the brain achieve this flexible motor control?
This question has been driving me from the beginning of my PhD thesis, and will continue to be central in my future research. Arm movements are an ideal model system for investigating skilled human actions. Thus, understanding their neural mechanisms informs universal motor control theories, and is fundamental for developing scientifically well-grounded learning and training regimes in applied areas such as clinical rehabilitation, sports, and biologically driven robotics. I adopt a systems approach to human motor control research, which links complex real-world behavior to its corresponding neural implementation. The central link is established via computational models, whose internal states relate to the neural processes and whose output corresponds to the behavior.
I am especially interested in the integration of visual and vestibular information for motor control. Regarding vision, I am primarily interested in the question: How does the brain bind actions to their visual consequences? This relationship is established by a dedicated visuo-motor binding mechanism, which I have started to investigate recently. My research on this topic offers a new perspective on fundamental aspects of visuo-motor control theories, and has the potential to better understand clinical conditions with disturbances in visuo-motor processing.

A dedicated binding mechanism for the visual control of movement

Attention has long been considered the main mechanism by which the brain filters sensory input for further processing. However, attention is a capacity-limited resource. This poses a problem for online control of goal-directed movements, during which the sensorimotor system must simultaneously extract visual information about the target(s) of the action, as well visual information about the hand(s). Moreover, it must do so in a complex visual world, additionally filtering out non-action related objects. Given these parallel demands on attentional resources, how does the motor system ensure robust and efficient feedback control?
We demonstrate that visual processing of hand motion employs a dedicated mechanism that circumvents the attentional bottleneck. In a series of experiments, we study the fast and automatic responses to visual displacements of the hand and target during reaching movements. We show that the processing of visual hand feedback is not influenced by the allocation of visual attention, whether endogenously or exogenously cued. In contrast, processing of visual target information is modulated by the locus of participants' attention. Furthermore, we demonstrate that the extraction of visual information about hand motion among multiple distracting objects is more efficient than the extraction of target motion. Together, these results provide strong evidence that the processing of re-afferent visual motion information can occur outside of the focus of visual attention, and is likely subserved by a distinct binding mechanism.
Related problems have been discussed in the cognitive and social neuroscience literatures under the term 'sense of agency' or 'self-other distinction'. We provide a radically different perspective to the problem of attributing visual information to one's own action by highlighting its importance for basic motor control. In the natural environment, the visual scene is cluttered with distracting objects. Visuomotor binding provides the necessary 'glue' between the sensory and motor systems, and therefore must constitute a central concept in theories of sensorimotor integration. Our research takes the first steps to characterise this mechanism.
This discovery raises questions regarding the computational and neural processes underlying the visuo-motor binding mechanism, and its relationship to visual attention and the sense of agency. I am currently modeling the influencing factors and computational processes leading to visuo-motor binding in a Bayesian estimation framework. This will establish a basic theoretical framework for visuo-motor binding.
Based on recent literature, I hypothesize that visuo-motor binding is disturbed in schizophrenic conditions with delusion of control. Thus, I aim to test schizophrenic patients in order to build a foundation for clinical research, which might in the long run contribute to improvements in early diagnosis of schizophrenia.

Computational and neural processes underlying the integration of visual hand and target information for feedback control of reaching movements

Visually guided movements rely on visual feedback from both the target and hand. The classical view on feedback control is that the brain extracts estimates of target and hand position from a visual scene, calculating a difference vector used by the motor system for online corrections. However, the complex computations required to calculate this difference vector could be too time consuming given the tight temporal constraints. Our research suggests that the fastest feedback corrections are generated by processing the two visual cues independently. This simpler computation seems to be better suited to the time demands for fast feedback control. In later phases of the movement correction, however, the responses of the motor system conform with a calculation of a difference vector. This means that the brain first computes a quick but sub-optimal solution to enable rapid response onset, which is later displaced by the sophisticated calculation that enables achievement of the task goal.
In a complementary fMRI study, we have started to determine tuning functions of single voxels with respect to these two visual sources of information. This will allow us to determine which of the possible models best characterizes the fine-grained neural activity pattern in the regions along the visuo-motor hierarchy. This project is exemplary for the tight connection between behavioral analysis, detailed computational modeling, and innovative analysis of neuroimaging data, that I consider essential for a comprehensive view on the neural processes underlying human motor control.

  Einen allgemeinverständlichen Artikel über meine Arbeit in London (2011-2014) finden Sie hier. Dort erkläre ich zwei speziellen Themen zur Verarbeitung visueller Informationen für die Bewegungssteuerung.

Redundancy in the motor system

I (co-)supervised a couple of graduate students on projects investigating how the brain solves the inherent redundancy in the motor system. A recently finished project sheds light on the mechanisms that assign visually detected errors to the involved limbs during a reaching movement (Reichenbach et al, 2013). The follow up experiments utilize a robotic manipulandum with 3 degrees of freedom, which enables us to separately measure and manipulate different segments of the reaching arm. We currently model the arm dynamics during natural reaching movements. Subsequently, we will quantitatively assess the changes in the dynamics during motor learning, which will form the basis for rehabilitation training regimes.

  Einen allgemeinverständlichen Artikel über die Arbeit meines medizinischen Doktoranden David Mehler finden Sie hier. Dort beschreibt er einen Beitrag der neurowissenschaftlichen Grundlagenforschung zur Rehabilitationsmedizin.

Research methods

psychophysics and behavioural testing with healthy human volunteers
virtual reality and robotic manipulanda
functional and anatomical MRI
neuro-navigated TMS
computational modelling

Research from my PhD thesis

Most of my past research investigated processing of sensory information for motor control. During my undergraduate studies, I started as a research assistant investigating multisensory integration. This work laid my foundation for adopting sophisticated experimental methods and introduced me to modeling neural processes in a Bayesian framework. The first behavioral study of my PhD thesis influenced the ongoing discussion about the representation of sensorimotor information in different reference frames (Reichenbach et al, 2009). My three subsequent TMS studies distinguished parietal areas that are crucial for the integration of visual, proprioceptive, and vestibular information into the current motor plan during reaching movements. These studies provide strong arguments relevant to the ongoing controversy about the human subregions involved in sensorimotor processes. For processing visual information, regarding both the target and the reaching arm, one key area is the left anterior IPS (Reichenbach et al, 2011). Interestingly, this area is traditionally regarded as 'grasping area'. Thus, my study challenges the strict functional division in the posterior parietal cortex (PPC), and suggests that this area is a critical node for visuo-motor integration in general. Processing proprioceptive information for reaching movements relies heavily on the posterior part of the left medial IPS, an area that is clearly distinct from multisensory areas (Reichenbach et al, in press). Finally, vestibular information seems to be processed predominantly in the right PPC (Reichenbach et al, in review). Moreover, I advanced methodological developments for investigating the specific cortical sub-areas that give rise to motor behavior in humans. By planning the TMS coil positions based on individual fMRI activations, I established a method to deal with the high inter-individual neuroanatomical differences in the human PPC, and improved the spatial resolution of event-related TMS.

  Einen allgemeinverständlichen Artikel über meine Doktorarbeit (2007-2010) finden Sie hier. Dort erläutere ich einige Grundlagen der neurowissenschaftlichen Sensomotorik Forschung und ihrer Methoden.