Research On Emotional Self-Regulation Based On Real-time FMRI Neurofeedback

Real-time functional Magnetic Resonance Imaging neurofeedback(rtfMRI-nf)has the advantage of higher spatial resolution and localization accuracy.Using rtfMRI-nf approaches,individuals can enhance their ability in cognition by receiving and regulating the fMRI signals from their brain activity in real-time.Facing the stress of daily life and work,the emotion regulation ability of much people will be disordered.So humans have paid much attention to how to use rtfMRI-nf enhancing people's emotion regulation ability.Emotion is a typical high level cognition function and the generation and regulation of it needs coordination and interaction of different functional regions.But the majority of current studies may be limited in terms of their ability to enable modulation of the entire neural circuitry involved in certain brain functions.So based on the analysis of brain regions and networks,the studies of how to use rtfMRI-nf training and assess training effect have important valuable of theory and clinical application.In this thesis,we explored some key issues of emotional regulation based on neural scientific and biomedicine research findings,including the classification of emotion states based on fMRI,the method of rtfMRI-nf emotion regulation based on multi-voxel pattern analysis,the method of assessment rtfMRI-nf training effect.The main works include:1.We investigated classification methods of brain emotion states,combining with neural scientific and anatomical medicine research findings about brain emotional cognition function.We compared the difference of selected voxels between univariate and multivariate methods by a designed emotion-related experiment.The results showed the multivariate method recursive feature elimination had a higher accuracy in emotion states recognition.Then we used this method to recognize emotion states in different emotion-related brain regions.The results indicated that multiple brain regions could provide more precise information with a higher accuracy.Those results laid a good foundation for rtfMRI-nf based on multiple brain regions.2.Majority of current studies may be limited in terms of their ability to enable modulation of the entire neural circuitry,which is not in accord with brain function.We proposed a rtfMRI-nf emotion regulation method based on multi-voxel pattern analysis which reflect the pattern changed in spatially distributed brain regions.The subjects could regulate their emotion states with the help of providing emotion states in real time.It has been found that participants were able to increase the separability of emotional states and enhance the activations of emotion networks,resulting mood improvement.The results indicated that this method is more sensitive to the pattern of emotion,which can reflect the central mechanisms of brain more effective.3.Based on local property of rsFC,we used the amygdala rsFC as a biomarker to assess training effect.To date,the majority of rtfMRI-nf studies have focused on the brain activation and behavior scales to demonstrate the effect of training.Resting-state functional connectivity(rsFC)is a highly effective and sensitive method for mapping complex neural circuits speculated to reflect the underlying neuroanatomy.Changes of rsFC after training are very important to assess the rtfMRI-nf training effect.It has been widely used in diagnosing and assessing of clinical diseases.The experiment results indicated that rtfMRI-nf training significantly increased rsFC of amygdala with prefrontal cortex and some others related to emotion,and exert the same effect on amygdala rsFC as the drug treatment effect in previous studies,which validiated the effectiveness of this method and provided a new method for assessing effect of training.4.Based on global property of rsFC,we used three core intrinsic connectivity networks(default mode network,central executive network,salience network)as a biomarker to assess the training effect.Brain network is the pyhsiological basis of information processing and cognition.The changes of brain network are more sensitive and effective in reflecting the function of emotion,so it has important valuable for assessing the rtfMRI-nf training effect.The experiment results indicated that rtfMRI-nf training could regulate the functional connectivity of core networks,improving abnormal functional connectivity induced by emotion regulation disorder,which validiated the effectiveness of this method and provided a new method for assessing effect of training.