Related Experiment Video
Updated: Jan 12, 2026
A Protocol for the Administration of Real-Time fMRI Neurofeedback Training
Published on: August 24, 2017
Modulating the sense of agency in functional neurological disorder using real-time fMRI neurofeedback: a
Eliane Müller1, Serafeim Loukas2, Salome Häuselmann3
1Department of Neurology, Faculty of Science and Medicine, University of Fribourg, Switzerland; Neuro-X Institute, Ecole Polytechnique Fédérale de Lausanne (EPFL), Geneva, Switzerland; Department of Neurology, Psychosomatic Medicine Unit, Inselspital Bern University Hospital, Switzerland; Graduate School of Cellular and Biomedical Sciences (GCB), University of Bern, Switzerland.
Background:
Disrupted sense of agency (SoA), the perceived loss of control over one's actions, is common in functional neurological disorder (FND) and linked to abnormal activation in agency-related networks, particularly the right temporoparietal junction (rTPJ). This proof-of-concept study tested whether fMRI neurofeedback (NF) could enhance SoA in FND by targeting the rTPJ.
Methods:
In this single-arm feasibility study, 18 patients with mixed FND symptoms completed three NF trainings during an adaptive visuomotor task, receiving intermittent visual feedback of their rTPJ activity. Pre- and post-training assessments included subjective agency ratings, neuroimaging during a transfer task, and clinical evaluations.
Results:
Group-level subjective agency significantly increased after training (p = 0.0083), driven by responders (n = 8) who exhibited rTPJ (p = 0.042) and supplementary motor area (SMA) upregulation (p = 0.015) in the first session, alongside consistent task engagement. Behavioral gains coincided with trend-level decreases in rTPJ-dorsolateral prefrontal cortex (dlPFC) connectivity (p = 0.098), while non-responders showed reduced rTPJ-SMA connectivity (p = 0.008). No group-level clinical improvements were observed, but higher initial rTPJ (p = 0.03, ρ = -0.78) and dlPFC (p = 0.03, ρ = -0.75) activation were negatively associated with changes in self-reported core symptoms.
Conclusion:
NF from rTPJ activity is feasible in FND and may enhance explicit SoA in a subset of patients through early neural modulation and sustained engagement. High inter-individual variability underscores the need for sham-controlled trials with optimized, engaging feedback designs and longitudinal monitoring to identify patients who could benefit from NF.
