Philosophy, Psychiatry, & Psychology

In their paper "Conceptual Challenges in the Neuroimaging of Psychiatric Disorders," Kanaan and McGuire (2011) review a number of methodological and analytical obstacles associated with the use of functional magnetic resonance imaging (fMRI) to study psychiatric disorders. Although we agree that there are challenges and limitations to this end, it would be a shame for those without a background in neuroimaging to walk away from this article with the impression that such work is too daunting, and thus not worth pursuing. On the contrary, despite a number of challenges (which are an inevitable part of all research), fMRI has already contributed many important insights into the nature and mechanisms of psychopathology and has the potential to contribute many more. Therefore, in the interest of providing an argument for the expanded use of fMRI in clinical research, we would like to briefly review the benefits of fMRI and how it can contribute to our understanding of psychiatric disorders.

Functional MRI has revolutionized the study of human thought, emotion, and behavior. Although lesion studies have provided and will continue to provide valuable insights into brain function, there are obvious limitations to its application that dramatically limit potential research populations and questions. As a noninvasive technology, fMRI gives us a relatively unique form of access to the human mind, in vivo.

As Kanaan and McGuire discussed, fMRI has a number of limitations, particularly when applied to psychiatric populations. Such limitations include the physical constraints of the scanning environment, the difficulties and complexities of trying to image "illness" per se, and the confounds of disorder heterogeneity and comorbidity. These are all important issues that certainly dictate (and thus limit) the tasks, study designs, and patient groups that realistically can be included in fMRI research. Thus, it is a fair conclusion that fMRI may not be the best tool with which to study all psychiatric disorders and all psychiatric research questions. Some disorders (e.g., anxiety, depression) are simply more conducive to fMRI research than others (e.g., attention deficit hyperactivity disorder, Tourette's syndrome), and some designs (e.g., group-based comparisons) are more appropriate than others (e.g., single-subject diagnoses).

Nevertheless, there are many important, clinically focused questions that can be uniquely addressed with functional neuroimaging. In addition [End Page 333] to allowing us to learn which neural structures are involved and impaired in psychopathology (i.e., the "where"), fMRI findings, when evaluated in the context of existing knowledge (as all results should be), can help us to understand which neural processes are involved and impaired (i.e., the "how" and "why"). As such, fMRI can ultimately help us to improve the assessment, prevention, and treatment of psychiatric illness. Although it is true that there are limited direct clinical applications for fMRI currently, it is essential for us to push forward with clinically based fMRI research to establish a solid foundation of knowledge from which many direct clinical applications can emerge in the not too distant future.

As the technology stands today, the most fruitful design approach for clinically focused fMRI research has two general characteristics. First, it involves the investigation of groups of subjects, rather than single individuals. This approach may involve comparing patients to healthy matched controls, comparing patients with disorder X with patients with disorder Y, or even examining individual differences within a group of patients with a common diagnosis. Second, instead of trying to capture episodes of "illness" in the scanner, fMRI tasks designed to index more basic cognitive, emotional, or behavioral processes that are predicted to be dysfunctional in certain disorders are used. This approach is similar to what Chris Frith has done in his research on schizophrenia, and obviates a dependence on the manifestation of patient symptoms during scanning sessions. As in all research, one must find an appropriate compromise between generalizability of the findings and rigorous control of the independent variables. In the earlier stages of a new field (such as fMRI), it is often more advantageous to lean toward rigorous control until the basic mechanisms are fairly well known. The power provided by...