Real-time fMRI and Neurofeedback of Brain Networks Mediating Trauma Memory Recall in PTSD

A support vector machine algorithm will be applied in real-time to fMRI data to identify distributed patterns of co-activated brain regions that specifically encode high emotional arousal (i.e,. high SCR) to the stress/trauma memory (note, this is equivalent to predictions of fitted Q-iteration in which the all actions are specified as zero, reward is equal to the support vector machine predicted arousal, and the discount factor of 0). The resulting idiosyncratic brain map would inform the neurofeedback phase in the next stage of fMRI data collection. This approach will first be piloted in the healthy participant group, then implemented in the PTSD participant group.