A Reckoning and Research Agenda for Neuroimaging in Psychiatry: Response to Henderson et al.

To the Editor: I thank Henderson et al. (1) for their interest in my review article in the Journal (2) and for their enthusiasm in bringing neuroimaging to real-world clinical care. Their point is mainly that additional, important forms of neuroimaging beyond functional MRI (fMRI) were not described in detail in my article. Although it is certainly true that there are many useful brain imaging tools available today, MRI-based research is presently the dominant modality in psychiatric neuroscience investigations. As such, as my primary goal was to examine the conceptual models guiding (or in some cases, misguiding) our research efforts and theoretical models of disease and treatment, I focused primarily on MRI. Indeed, regardless of method, my central thesis is that it is the broader conceptual framework in which these methods are used that is most problematic.

Likewise, I agree with Henderson et al. that consideration of ultimate clinical translation is critical. To that effect, I briefly highlighted the important role that EEG can play. For example, we have recently found that a particular EEG connectivity method can reveal resting-state fMRI-like network structure (3), distinguish response to an antidepressant medication and placebo when assessed prior to treatment (4), and provide rich data to support robust individual-level treatment prediction through machine learning (5). Indeed, given the use of radioactivity and the higher cost (relative to EEG) of positron emission tomography or single-photon emission computed tomography, as advocated for by Henderson et al., EEG may ultimately prove to be the ideal tool for real-world applications. Nonetheless, unless the way we ask and answer questions using any form of neuroimaging changes, it is my position that little is likely to change in terms of either mechanistic insight or creation of diagnostics ready for clinical care.