@phdthesis{Tupak2013,
  author    = {Tupak, Sara},
  title     = {Modulators of Prefrontal Fear Network Function: An Integrative View},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-85673},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2013},
  abstract  = {Regulating our immediate feelings, needs, and urges is a task that we are faced with every day in our lives. The effective regulation of our emotions enables us to adapt to society, to deal with our environment, and to achieve long-term goals. Deficient emotion regulation, in contrast, is a common characteristic of many psychiatric and neurological conditions. Particularly anxiety disorders and subclinical states of increased anxiety are characterized by a range of behavioral, autonomic, and neural alterations impeding the efficient down-regulation of acute fear. Established fear network models propose a downstream prefrontal-amygdala circuit for the control of fear reactions but recent research has shown that there are a range of factors acting on this network. The specific prefrontal cortical networks involved in effective regulation and potential mediators and modulators are still a subject of ongoing research in both the animal and human model. The present research focused on the particular role of different prefrontal cortical regions during the processing of fear-relevant stimuli in healthy subjects. It is based on four studies, three of them investigating a different potential modulator of prefrontal top-down function and one directly challenging prefrontal regulatory processes. Summarizing the results of all four studies, it was shown that prefrontal functioning is linked to individual differences in state anxiety, autonomic flexibility, and genetic predisposition. The T risk allele of the neuropeptide S receptor gene, a recently suggested candidate gene for pathologically elevated anxiety, for instance, was associated with decreased prefrontal cortex activation to particularly fear-relevant stimuli. Furthermore, the way of processing has been found to crucially determine if regulatory processes are engaged at all and it was shown that anxious individuals display generally reduced prefrontal activation but may engage in regulatory processes earlier than non-anxious subjects. However, active manipulation of prefrontal functioning in healthy subjects did not lead to the typical behavioral and neural patterns observed in anxiety disorder patients suggesting that other subcortical or prefrontal structures can compensate for an activation loss in one specific region. Taken together, the current studies support prevailing theories of the central role of the prefrontal cortex for regulatory processes in response to fear-eliciting stimuli but point out that there are a range of both individual differences and peculiarities in experimental design that impact on or may even mask potential effects in neuroimaging research on fear regulation.},
  subject      = {Neurogenetik},
  language  = {en}
}
@phdthesis{Kaufmann2013,
  author    = {Kaufmann, Tobias},
  title     = {Brain-computer interfaces based on event-related potentials: toward fast, reliable and easy-to-use communication systems for people with neurodegenerative disease},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-83441},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2013},
  abstract  = {Objective: Brain Computer Interfaces (BCI) provide a muscle independent interaction channel making them particularly valuable for individuals with severe motor impairment. Thus, different BCI systems and applications have been proposed as assistive technology (AT) solutions for such patients. The most prominent system for communication utilizes event-related potentials (ERP) obtained from the electroencephalogram (EEG) to allow for communication on a character-by-character basis. Yet in their current state of technology, daily life use cases of such systems are rare. In addition to the high EEG preparation effort, one of the main reasons is the low information throughput compared to other existing AT solutions. Furthermore, when testing BCI systems in patients, a performance drop is usually observed compared to healthy users. Patients often display a low signal-to-noise ratio of the recorded EEG and detection of brain responses may be aggravated due to internally (e.g. spasm) or externally induced artifacts (e.g. from ventilation devices). Consequently, practical BCI systems need to cope with mani-fold inter-individual differences. Whilst these high demands lead to increasing complexity of the technology, daily life use of BCI systems requires straightforward setup including an easy-to-use graphical user interface that nonprofessionals can handle without expert support. Research questions of this thesis: This dissertation project aimed at bringing forward BCI technology toward a possible integration into end-users' daily life. Four basic research questions were addressed: (1) Can we identify performance predictors so that we can provide users with individual BCI solutions without the need of multiple, demanding testing sessions? (2) Can we provide complex BCI technology in an automated, user-friendly and easy-to-use manner, so that BCIs can be used without expert support at end-users' homes? (3) How can we account for and improve the low information transfer rates as compared to other existing assistive technology solutions? (4) How can we prevent the performance drop often seen when bringing BCI technology that was tested in healthy users to those with severe motor impairment? Results and discussion: (1) Heart rate variability (HRV) as an index of inhibitory control (i.e. the ability to allocate attention resources and inhibit distracting stimuli) was significantly related to ERP-BCI performance and accounted for almost 26\% of variance. HRV is easy to assess from short heartbeat recordings and may thus serve as a performance predictor for ERP-BCIs. Due to missing software solutions for appropriate processing of artifacts in heartbeat data (electrocardiogram and inter-beat interval data), our own tool was developed that is available free of charge. To date, more than 100 researchers worldwide have requested the tool. Recently, a new version was developed and released together with a website (www.artiifact.de). (2) Furthermore, a study of this thesis demonstrated that BCI technology can be incorporated into easy-to-use software, including auto-calibration and predictive text entry. Na{\"i}ve, healthy nonprofessionals were able to control the software without expert support and successfully spelled words using the auto-calibrated BCI. They reported that software handling was straightforward and that they would be able to explain the system to others. However, future research is required to study transfer of the results to patient samples. (3) The commonly used ERP-BCI paradigm was significantly improved. Instead of simply highlighting visually displayed characters as is usually done, pictures of famous faces were used as stimulus material. As a result, specific brain potentials involved in face recognition and face processing were elicited. The event-related EEG thus displayed an increased signal-to-noise ratio, which facilitated the detection of ERPs extremely well. Consequently, BCI performance was significantly increased. (4) The good results of this new face-flashing paradigm achieved with healthy participants transferred well to users with neurodegenerative disease. Using a face paradigm boosted information throughput. Importantly, two users who were highly inefficient with the commonly used paradigm displayed high accuracy when exposed to the face paradigm. The increased signal-to-noise ratio of the recorded EEG thus helped them to overcome their BCI inefficiency. Significance: The presented work at hand (1) successfully identified a physiological predictor of ERP-BCI performance, (2) proved the technology ready to be operated by na{\"i}ve nonprofessionals without expert support, (3) significantly improved the commonly used spelling paradigm and (4) thereby displayed a way to effectively prevent BCI inefficiency in patients with neurodegenerative disease. Additionally, missing software solutions for appropriate handling of artifacts in heartbeat data encouraged development of our own software tool that is available to the research community free of charge. In sum, this thesis significantly improved current BCI technology and enhanced our understanding of physiological correlates of BCI performance.},
  subject      = {Gehirn-Computer-Schnittstelle},
  language  = {en}
}