Advances in real-time fMRI neurofeedback - from big data to individualized clinical applications

Real-time functional Magnetic Resonance Imaging (fMRI) neurofeedback is a neuroimaging technique that allows individuals to gain voluntary control over their own brain signals. Previous real-time fMRI neurofeedback tudies have demonstrated that successful regulation of predefined brain regions of interest can lead to improvements in behavior in healthy individuals and improvements in clinical measures in a wide range of different patient populations. To further advance these applications of neurofeedback training, much progress has been made in recent years with regards to the development of novel neurofeedback methods and the tailoring of experimental paradigms to specific applications and patient groups. At the same time, the field of neurofeedback has also matured in that more and more data have been collected to aggregate information and to eventually gain a better understanding of which factors determine successful neurofeedback training. Both, novel neurofeedback approaches and the identification of generalizable factors that influence neurofeedback success can help to improve neurofeedback efficacy. In this Thesis, I pursued both of these complementary approaches to further advance the field of neurofeedback: First, I established a novel neurofeedback intervention that made use of an innovative brain-based adaptive closed-loop paradigm. I applied this new neurofeedback intervention to individuals with nicotine use disorder to support their smoking cessation. Second, I employed big data meta-analyses and machine learning methods to investigate factors that influence neurofeedback success across a large database of neurofeedback studies that I compiled. The results from this Thesis indicate that adaptive closed-loop neurofeedback can help individuals with nicotine use disorder to reduce nicotine dependence and consumed cigarettes. They further demonstrate that big data meta-analyses and machine learning algorithms are promising approaches for the determination of factors that influence neurofeedback success, however more data will be needed in the future. Thus, this Thesis might contribute to establishing a new form of neurofeedback intervention for substance use disorder as well as other clinical conditions, and to improving the efficacy of neurofeedback as a promising scientific tool and novel brain-based intervention