Fig. 1: Experimental paradigm.

A Diagram of the fMRI experiment. On average participants (N = 20) completed the experiment in around two weeks, with a minimum break of one day between each session (neurofeedback: NF). B One block of neurofeedback training is shown. In total six were done per run. Participants start at rest. After one auditory cue the baseline condition starts and participants move the robot in the synchronous mode, while listening to the minimal level of auditory feedback (it did not vary with brain activity at this point). Participants were told they shouldn’t engage in a mental strategy here and should rather “clear their minds”. After two auditory cues the feedback condition started, and participants continued moving the robot which changed to the asynchronous mode. The auditory feedback (resembling a looming and receding wave) became now variable with the spatial correlation between ongoing brain activity and the PH-network (visually represented three times in the figure but it varied every 1.5 s). Participants knew they should deploy a mental strategy to boost the feedback to the maximum level. This condition ended with three auditory cues. Participants were asked about preferred strategies at the end of each session (and reassured some strategies might not be verbalizable). At the end of the last training sessions a more general briefing occurred. C Two blocks of the riPH task are shown in sequence of a total of 16 blocks per run. Blocks were composed of rest followed by an auditory cue which indicated the start of the condition were participants moved the robot in asynchronous or synchronous mode. Two auditory cues indicated the end of the riPH task. No more than two blocks in sequence could be of the same condition. At the end of the second run of the riPH task, participants performed two individual blocks of robot manipulation (one asynchronous, one synchronous), each followed by a questionnaire assessing PH induction sensitivity amongst other sensations and control questions.