Self-Regulation of Blood Oxygenation Level Dependent Response: Primary Effect or Epiphenomenon?

In the last decade, several studies indicated that neuronal activity can be volitionally modulated using real-time fMRI (rtfMRI) based neurofeedback. Human participants through rtfMRI paradigms can learn to regulate the blood oxygenation level dependent (BOLD) response in several localized cortical and subcortical regions (for extended reviews see Caria et al., 2012; Weiskopf, 2012; Sulzer et al., 2013). Increasing evidence also indicated that strengthening or weakening specific BOLD activity using rtfMRI training leads to significant changes in cognitive, emotional and motor behavior (Caria et al., 2012; Weiskopf, 2012; Sulzer et al., 2013). These findings suggested that rtfMRI might represent an important novel approach in cognitive neuroscience by potentially providing indications of cause-and-effect relationships between brain and behavior, and also in clinical applications (Subramanian et al., 2011; Linden et al., 2012; Ruiz et al., 2013; Sitaram et al., 2014). Although rtfMRI studies extend and enrich a large body of literature demonstrating operant learning of neuronal oscillations, skepticism still exists on the validity and reliability of studies showing learned control of the BOLD response. In particular, it is still debated whether this phenomenon is a primary effect of learning or it just an epiphenomenon resulting, for instance, from repeated execution of some mental processes. Here, I will discuss specific psychophysiological and neurophysiological mechanisms presumably underlying learned regulation of the BOLD response to attempt to clarify its nature.