Targeting large-scale networks in depression with real-time fMRI neurofeedback

点击翻译按钮获取中文摘要

Abstract Major depressive disorder (MDD) is a common, debilitating illness, and new treatments are desperately needed. Addressing this critical need, this exploratory R21 proposal will use real-time fMRI neurofeedback (rtfMRI-NF), which has demonstrated that individuals can self-modulate brain activity. Training patients with depression with rtfMRI-NF paradigms has demonstrated promising, but modest, improvement in depressive symptoms, and new refinements in targeting key brain networks should generate better therapeutic leverage. fMRI studies of large-scale networks (LSN) in the brain have identified aberrant connectivity within and between networks in MDD, as well as other psychiatric disorders. Accordingly, this proposal will develop a new rtfMRI-NF paradigm that specifically targets interactions between two large-scale networks, critical for healthy psychological functioning – the salience network (SN) and the default-mode network (DMN). A fundamental observation of LSNs is that the DMN, activated by internally generated content, deactivates during tasks requiring attention to external stimuli. The SN, along with other ‘task positive’ networks, exhibit a reciprocal relationship with the DMN, becoming active in tasks requiring external direction of attention and executive functions when the DMN deactivates. Nodes within the SN, such as the anterior insula, have been suggested to mediate the balance between the DMN and task positive networks. As imbalanced connectivity amongst these networks is linked to depression, methods to boost SN function may improve therapeutic responses. For the first step in a larger program to test this hypothesis, this proposal will develop a ’Recall Stop Task’ (RST), using rtfMRI-NF, to target “switching off” the DMN and activating the SN. Forty patients with active MDD will be randomized in a double-blinded, controlled study. In Aim 1, during an initial (localizer) MRI session, network participation in this network switching task will be evaluated, testing the hypothesis that SN is engaged by the RST. For Aim 2, participants will be randomized to receive either valid NF from the personalized network activated by the switching task during the localizer session, or they will receive sham NF. We will test the hypothesis that real, compared to sham NF, will increase activation in the RST, and that the NF task will increase SN connectivity. Successful development of this switching network NF paradigm will provide a robust, neuroscience-informed target for personalized interventions designed to engage networks at the level of the individual subject. The next step would entail an R61/R33 project to demonstrate target engagement and test the hypothesis that symptom amelioration for MDD will follow. Although subsequent steps will focus on MDD, benefits are predicted to be transdiagnostic.