Real-time fMRI Neurofeedback
Schematic representation of the real-time fMRI (rt-fMRI) neurofeedback experimental procedure.
A) Baseline resting state (RS-pre) scan. B) Functional Localization of the default mode network (DMN) and the central executive network (CEN). C) rt-fMRI feedback process showing the online monitoring of brain states while performing “mental noting”. If rt-fMRI analysis resulted in a Positive Diametric Activity (PDA) score (red shadowing) the central white dot of the feedback display moved in the upper direction and towards the red circle. The contrary was the case when a negative PDA score was triggered (blue shadowing). D) Post rt-fMRI resting state.
Real Time fMRI Feedback Targeting Default Mode Network (DMN) Reduces Auditory Hallucinations
Susan Whitfield-Gabrieli, Clemens Bauer, Kana Okano, Paul Nestor, Elisabetta Del Re, Satra Gosh, Margaret Niznikiewicz (2017) Schizophrenia Bulletin, 43(suppl 1):S233,https://doi.org/10.1093/schbul/sbx022.059
Background: Neuroimaging studies have consistently found hyperconnectivity of the DMN as well as reduced anticorrelations between DMN and Central Executive Network (CEN) in patients with schizophrenia (SZ). This may reflect an inability to redirect resources away from internal thoughts and feelings towards external stimuli, leading to disorder related symptomatology. Moreover, greater DMN functional connectivity correlates with auditory hallucinations (AH) and reduced anticorrelations are associated with greater cognitive deficits. Noting practice (an essential part of Mindfulness training [MT]) has been shown to decrease DMN, engage CEN, and to ameliorate clinical symptoms of SZ. We show evidence that SZ patients can learn to increase the positive diametric activity (PDA) defined as increased CEN and decreased DMN activity through fMRI neurofeedback-enhanced MT from these networks.
Methods: Nine schizophrenia patients underwent (1) two 6-minute resting-state (RS) scans, (2) 2 no-feedback transfer task (TR) scans, and (3) 4 feedback scans. RS-1 was used to extract the DMN & CEN networks. All processing was performed in FSL build 5.08. ICA was performed and the components correlated to the DMN/CEN of Yeo et al 2011, thresholded and binarized. rt-fMRI-neurofeedback: sessions consisted of noting practice (MT) while receiving feedback from DMN/CEN activity with BOLD fluctuations measured using rt-fMRI. Participants were instructed to move a white dot into an upper-yellow circle by performing the noting practice. Connectivity analysis was performed using the Conn Toolbox.
Results: All patients could (1) increase PDA by performing mental labeling/noting practice while receiving rt-fMRI feedback (2) increase DMN/CEN anticorrelations (MPFC and right DLPFC), and (3) reduce intrinsic positive DMN connectivity (MPFC/PCC). Finally, AH were significantly reduced post-rt-fMRI.
Conclusion: These results demonstrate that SZ patients can modulate PDA by performing MT and that neurofeedback-enhanced MT produces an increase in anticorrelation of MPFC/rDLPFC connectivity and a reduction of intrinsic positive DMN connectivity. Moreover, these patients had a subsequent reduction in AH. Both intrinsic DMN and DMN/CEN anticorrelations provide targets to study the neurobiology of SZ symptomatology, especially AH, and may aid the development of novel forms of interventions aiming to “normalize” the brain’s resting state.
- Stoeckel LE, Garrison KA, Ghosh S, Wighton P, Hanlon CA, Gilman JM, Greer S, Turk-Browne NB, deBettencourt MT, Scheinost D, Craddock C, Thompson T, Calderon V, Bauer CC, George M, Breiter HC, Whitfield-Gabrieli S, Gabrieli JD, LaConte SM, Hirshberg L, Brewer JA, Hampson M, Van Der Kouwe A, Mackey S, Evins AE. (2014) Optimizing real time fMRI neurofeedback for therapeutic discovery and development. Neuroimage Clin.; 5:245-55.
- Hinds O, Thompson TW, Ghosh S, Yoo JJ, Whitfield-Gabrieli S, Triantafyllou C, Gabrieli JD. (2013) Roles of default-mode network and supplementary motor area in human vigilance performance: evidence from real-time fMRI. J Neurophysiol.; 109(5):1250-8.
- Hinds O, Ghosh S, Thompson TW, Yoo JJ, Whitfield-Gabrieli S, Triantafyllou C, Gabrieli JD. (2011) Computing moment-to-moment BOLD activation for real-time neurofeedback. Neuroimage; 54(1):361-8.
- deCharms RC, Christoff K, Glover GH, Pauly JM, Whitfield S, Gabrieli JD. (2004) Learned regulation of spatially localized brain activation using real-time fMRI. Neuroimage; 21(1):436-43.