Self-regulation strategy, feedback timing and hemodynamic properties modulate learning in a simulated fMRI neurofeedback environment

Abstract

Direct manipulation of brain activity can be used to investigate causal brain-behavior relationships. Current noninvasive neural stimulation techniques are too coarse to manipulate behaviors that correlate with fine-grained spatial patterns recorded by fMRI. However, these activity patterns can be manipulated by having people learn to self-regulate their own recorded neural activity. This technique, known as fMRI neurofeedback, faces challenges as many participants are unable to self-regulate. The causes of this non-responder effect are not well understood due to the cost and complexity of such investigation in the MRI scanner. Here, we investigated the temporal dynamics of the hemodynamic response measured by fMRI as a potential cause of the non-responder effect. Learning to self-regulate the hemodynamic response involves a difficult temporal credit-assignment problem because this signal is both delayed and blurred over time. Two factors critical to this problem are the prescribed self-regulation strategy (cognitive or automatic) and feedback timing (continuous or intermittent). Here, we sought to evaluate how these factors interact with the temporal dynamics of fMRI without using the MRI scanner. We first examined the role of cognitive strategies by having participants learn to regulate a simulated neurofeedback signal using a unidimensional strategy: pressing one of two buttons to rotate a visual grating that stimulates a model of visual cortex. Under these conditions, continuous feedback led to faster regulation compared to intermittent feedback. Yet, since many neurofeedback studies prescribe implicit self-regulation strategies, we created a computational model of automatic reward-based learning to examine whether this result held true for automatic processing. When feedback was delayed and blurred based on the hemodynamics of fMRI, this model learned more reliably from intermittent feedback compared to continuous feedback. These results suggest that different self-regulation mechanisms prefer different feedback timings, and that these factors can be effectively explored and optimized via simulation prior to deployment in the MRI scanner.

Advances in fMRI Real-Time Neurofeedback

“…There are in general two ways regarding how often feedback scores are presented during the fMRI neurofeedback training; continuously and intermittently [50–52]. Here, continuous feedback refers to the feedback that is presented continuously during the neurofeedback training, and the feedback score changes every time the brain activation is measured by fMRI, for instance, every 2 sec.…”

“…It is controversial which type of feedback is more effective. A recent computational study [52] has indicated that for successful fMRI neurofeedback learning, temporal characteristics of fMRI signal need to be considered since the fMRI signal is delayed and blurred in time relative to underlying neural activities due to slower hemodynamics than neuronal firing. These characteristics form a temporal credit assignment problem [53] in which participants need to associate their neural activities at a certain time point with a feedback score that reflects the delayed and blurred fMRI signals.…”

“…These characteristics form a temporal credit assignment problem [53] in which participants need to associate their neural activities at a certain time point with a feedback score that reflects the delayed and blurred fMRI signals. The results of this study [52] suggest that the continuous feedback is more effective when participants are provided with explicit knowledge of how the feedback is calculated since this knowledge helps participants solve the temporal credit assignment problem. On the other hand, intermittent feedback may lead to better neurofeedback learning without such knowledge (i.e., implicit neurofeedback training) since the intermittent feedback largely reduces the temporal credit assignment problem.…”

Control freaks: Towards optimal selection of control conditions for fMRI neurofeedback studies

“…The "modelling" phase can involve experimental brain-imaging studies and comparison of different neurofeedback protocols in healthy participants or patients, but also exclusively rely on modelled data obtained from simulations. It has recently been shown that simulation studies can be useful for the assessment of feasibility of real-time fMRI neurofeedback protocols (5). The next stage would be an "exploratory trial", which could employ an adaptive design, in order to optimize the intervention and assess its feasibility in the targeted patient population.…”

Multivoxel Neural Reinforcement Changes Resting-State Functional Connectivity Within the Threat Regulation Network

“…Admittedly, neural reinforcement is still a new intervention and few studies on its mechanisms of action are available (see (Oblak et al, 2017;Shibata et al, 2019) . Our current results suggest the interesting possibility that the changes in the regulatory action of the vlPFC and vmPFC might have led to the decrease in amygdala and skin conductance reactivity.…”