Functional MRI neurofeedback (NF) allows humans to self-modulate neural patterns in specific brain areas. This technique is regarded as a promising tool to translate neuroscientific knowledge into brain-guided psychiatric interventions. However, its clinical implementation is restricted by unstandardized methodological practices, by clinical definitions that are poorly grounded in neurobiology, and by lack of a unifying framework that dictates experimental choices. Here we put forward a new framework, termed 'process-based NF', which endorses a process-oriented characterization of mental dysfunctions to form precise and effective psychiatric treatments. This framework relies on targeting specific dysfunctional mental processes by modifying their underlying neural mechanisms and on applying process-specific contextual feedback interfaces. Finally, process-based NF offers designs and a control condition that address the methodological shortcomings of current approaches, thus paving the way for a precise and personalized neuromodulation. The use of functional MRI (fMRI) in neurofeedback (fMRI-NF) has brought new hope to the field of self-guided neuromodulation. fMRI-NF allows individuals to modulate spatially localized neural patterns in real-time, using contingent rewarding feedback. Accumulating evidence suggests that in many cases, attaining significant neural modulations in line with the task protocol (i.e., NF success) is followed by corresponding mental and behavioural changes1, thus contributing to bridging the gap between brain functionality and our mental experience. Despite this promising prospect, the utilization of fMRI-NF for basic science as well as for clinical purposes has been slower than expected. This may be due to various methodological constraints, such as the lack of proper control conditions and inadequate blinding and randomization, as well as the relatively small sample sizes that characterize the field. Furthermore, brain-guided interventions do not correspond with current psychiatric categorization, which traditionally relies on subjective reports rather than on

Highlights Randomized clinical trial with a novel self-neuromodulation training in PTSD. Demonstration of feasibility of an fMRI-informed EEG model of Amygdala modulation (AmygEFP). Individually-tailored trauma-related content as the training feedback interface. Results showed reduction of PTSD symptoms following AmygEFP trauma-related feedback training.

We present a deep neural network method for learning a personal representation for individuals that are performing a self neuromodulation task, guided by functional MRI (fMRI). This neurofeedback task (watch vs. regulate) provides the subjects with a continuous feedback contingent on down regulation of their Amygdala signal and the learning algorithm focuses on this region's time-course of activity. The representation is learned by a self-supervised recurrent neural network, that predicts the Amygdala activity in the next fMRI frame given recent fMRI frames and is conditioned on the learned individual representation. It is shown that the individuals' representation improves the next-frame prediction considerably. Moreover, this personal representation, learned solely from fMRI images, yields good performance in linear prediction of psychiatric traits, which is better than performing such a prediction based on clinical data and personality tests. Our code is attached as supplementary and the data would be shared subject to ethical approvals.

The reciprocal connections between emotion and attention are vital for adaptive behaviour. Previous results demonstrated that the behavioural effects of emotional stimuli on performance are attenuated when executive control is recruited. The current research studied whether this attenuation is modality dependent. In two experiments, negative and neutral pictures were presented shortly before a visual, tactile, or auditory target in a Simon task. All three modalities demonstrated a Simon effect, a conflict adaptation effect, and an emotional interference effect. However, the interaction between picture valence and Simon congruency was found only in the visual task. Specifically, when the Simon target was visual, emotional interference was reduced during incongruent compared to congruent trials. These findings suggest that although the control-related effects observed in the Simon tasks are not modality dependent, the link between emotion and executive control is modality dependent. Presumably, this link occurs only when the emotional stimulus and the target are presented in the same modality.

No abstract