Stroke is caused when an artery carrying blood from heart to an area in the brain bursts or a clot obstructs the blood flow thereby preventing delivery of oxygen and nutrients. About half of the stroke survivors are left with some degree of disability where the impairment of motor control has been mentioned most frequently as the most important disability. Therefore, innovative methodologies for stroke neurorehabilitation are urgently required to reduce long-term disability.
Neuroplasticity is the ability of the central nervous system to respond to intrinsic or extrinsic stimuli by reorganizing its structure, function and connections. Neuroplasticity is involved in post-stroke restorative rehabilitation of upper-limb function, but also can cause maladaptive functional outcomes, which can compromise re-gain of function via implementation of sub-optimal compensatory movement strategies. Such neuroplastic changes can be facilitated with noninvasive brain stimulation (NIBS) techniques, such as transcranial direct current stimulation (tDCS). tDCS - an electrically based intervention directed at the central nervous system level - is a promising tool to facilitate neuroplasticity in stroke rehabilitation. In this work, we investigate on a physiological signal changes appearing on functional near-infrared spectroscopy (fNIRS) and electroencephalography (EEG) neuroimaging systems to objectively quantify the progress of a chosen tDCS treatment regime, correlating outcome with brain activation patterns as a marker of the underlying neuronal plasticity. Here, it was postulated that tDCS (Dietzel and Heinemann 1986) that perturbs hemodynamic (fNIRS) and electrophysiological (EEG) responses where the interactions between the hemodynamic and electrophysiological responses, captured with NIRS-EEG joint modeling, may provide an assessment of neurovascular coupling. Such an approach is novel since it introduces neuroimaging in a field so far not accessible by existing technology. We use adaptive identification techniques to correspond to the time-variances as previously applied to other application for muscular response changes.
Title: Computational modeling for brain dynamics time-variance evoked by noninvasive brain stimulation. Keywords: Computational modeling, Neuroplasticity, Adaptive tracking, Brain dynamics, Signal processing
BrainStateDecoding Lab of Dr. Michael Tangermann (Computer Science Dept.); payment level: TV-L E13 (65%); duration: 3/2015-2/2017
Brain-Computer Interface (BCI) systems make use of machine learning methods to cope with the challenges which appear during the decoding of individual brain signals in real-time. Within the BrainLinks-BrainTools cluster of excellence at the University of Freiburg (Germany), we evaluate whether BCI methods can be applied clinically outside of classical control paradigms.
The PhD student will explore the use of BCI methods to support neurological rehabilitation training of speech and attention-related deficits after stroke.
Scientific challenges comprise the development and investigation of brain signal analysis methods capable to describe rapid fluctuations of brain states and to assess the dynamics of network structures connecting brain areas.
Working Areas (among others):
- Theories (statistics, mathematics) and algorithms in the field of machine learning, with special emphasis on adaptive methods for the real-time decoding of mental states; software implementations thereof.
- Conducting EEG studies with healthy controls and stroke patients.
- Analysis of EEG experiments.
- Supervision of Bachelor students.
- Scientific dissemination of results (conferences, publications).
- Excellent Master studies on Computer Science, Mathematics, Electrical Engineering, Biomedical Engineering, Cognitive Science or closely related fields.
- Knowledge in theory and methods of machine learning / artificial intelligence; very good math knowledge (specifically in probability theory, statistics, linear algebra).
- Good knowledge about the design, the analysis and implementation of algorithms; experience with mathematical software (e.g. Matlab).
- Hands-on experience in the design, execution and analysis of electrophysiological experiments, and in signal processing of EEG-, EMG-, EOG- or ECoG data.
- Excellent communication skills in English
The BrainStateDecoding Lab:
The international lab is embedded into the Cluster of Excellence BrainLinks-BrainTools at the University of Freiburg. The cluster provides supporting career actions for PhD students. Tight collaborations with other BLBT groups are the basis for investigating novel machine learning approaches for the real-time analysis of brain signals. The activities range from theory development to BCI applications in a clinical context and for healthy users. Implementing an equal-chances policy and family-friendly working conditions, we explicitly encourage applications of female researchers. Handicapped applicants will be given priority to non-handicapped applicants, if they have comparable qualifications. Further information: www.bsdlab.uni-freiburg.de, www.brainlinks-braintools.uni-freiburg.de and via Dr. Tangermann.
How to apply: