Preview

PDF, English - main document Download (5MB) | Terms of use

Abstract

Psychiatric disorders, in particular schizophrenia and bipolar disorder, affect the patients’ lives deeply on many levels and place a heavy burden on the healthcare system. The treatment of these diseases is often complicated and marked by many setbacks. Symptoms that have the strongest consequences for coping with everyday life are the impairments of cognitive performance, for example memory or attention deficits. Therefore, it is of great interest to better understand the underlying pathomechanisms to eventually improve treatment options for those patients. In this thesis two different fMRI studies were used to investigate the functional correlates of patients suffering from schizophrenia or bipolar disorder while performing a combined oddball-incongruence task and a reward associated task.

Study A conducted a categorical comparison between bipolar and schizophrenia patients of the brain activation during an oddball and incongruence task. The results showed pathophysiological differences in the activation intensities between bipolar and schizophrenia patients as well as between the patient groups and healthy individuals. Overall it seems as if the brain activation severely depended on the task difficulty leading to compensatory hyperactivations in frontal brain areas of bipolar patients during the oddball task. Schizophrenia patients demonstrated low threshold hyperactivations in the intraparietal cortex compared to healthy controls. In the cognitively more demanding incongruence condition these compensatory mechanisms seemed to fail leading to hypoactivations in various brain regions such as the middle frontal gyrus or ventral pathway.

Pilot study B searched retrospectively for functional markers which enable support vector machine algorithms predicting specific treatment response to typical and atypical antipsychotics as well as aripiprazole in a transnosological sample consisting of bipolar and schizophrenia patients. Consequently, (de-)activation differences between responders and non-responders in their respective treatment arm resulting from the desire-reason-dilemma paradigm were applied to support vector machine algorithms. The implementation of parameter estimates from deactivations of aripiprazole non-responders in brain regions partially associated with the default mode network, led to a successful treatment response prediction of patients receiving aripiprazole.

Even though in future studies the sample sizes should be increased and monotherapeutical treatment ensured, this thesis already provides important insights on the pathomechanisms of bipolar disorder and schizophrenia patients or more specifically within the spectrum of both diseases. Prospectively, further studies can help to specify potential functional biomarkers which also might be able to predict treatment response and consequently approach personalized precision treatment in psychiatric disorders.