Secondary Logo

Neuroimaging in Psychiatry: A Quarter Century of Progress

Silbersweig, David A. MD*; Rauch, Scott L. MD

doi: 10.1097/HRP.0000000000000177
Advances in Psychiatric Research and Practice 25th Anniversary Brief Communication
Free

*Harvard Medical School and Department of Psychiatry, Brigham and Women’s Hospital, Boston, MA. Email: dsilbersweig@bwh.harvard.edu

Harvard Medical School and McLean Hospital, Belmont, MA. Email: srauch@partners.org

In many respects, advances in neuroimaging have transformed the field of psychiatry over the past quarter century. The field has yet to deliver, however, on the promise of translating these advances to substantially improved care or better outcomes.

The ability to visualize and quantify brain structure and function—in vivo, noninvasively—is particularly important for psychiatric illnesses. These conditions produce disruptions of perception, cognition, emotion, and behavior that affect distinctly human mental life and behavior. Such phenomena have long been difficult to associate with biological substrates, resulting in residual dualism and stigma. The absence of macroscopic neuropathology has also separated psychiatric from neurologic disorders: the identification of a stroke or tumor with neuroimaging confirms disease localization and guides clinical decision making, and post-mortem findings help to define disorders more clearly by their demonstrable pathology in the brain. While contemporary psychiatric neuroimaging techniques have yet to deliver clinical utility, they have permitted an increasingly sophisticated understanding of brain circuit abnormalities in psychiatric disease. This increased understanding, in combination with advances in basic neuroscience, has laid a foundation for medical progress.

The late 1980s set the stage for the functional-imaging revolution, building upon earlier efforts that established structural tomographic imaging via CT. A few major academic medical centers had developed interdisciplinary positron emission tomography groups and facilities (PET, which required cyclotrons, was more precise than SPECT, single-photon emission computed tomography, which could be performed in nuclear medicine departments). 18-fluorodeoxyglucose (FDG) PET, measuring glucose metabolism associated with neuronal activity, provided a robust baseline map of resting activity, and allowed a number of important comparisons among patient and control groups. Shortly thereafter, H2(15)O PET, with multiple measures of regional cerebral blood flow (also localized to neuronal activity), allowed the isolation of differentially active brain states associated with particular mental states; what enabled this advance was the development of neuropsychological (aka cognitive-affective neuroscience) activation paradigms with control conditions, performed in the scanner.1

Neural circuit dysfunction associated with psychiatric symptoms or syndromes could now be localized. A number of fundamental findings were produced with these methods, including demonstrations of dorsolateral prefrontal hypoactivity in schizophrenia,2 subgenual prefrontal cortex hyperactivity in major depression,3 neural correlates of psychotic hallucinations,4 cortico-striatal dysfunction in OCD and related disorders,5 and amygdalar hyperreactivity in anxiety disorders and PTSD.6 Such studies established typical profiles of increased limbic emotional processing in the presence of decreased prefrontal executive processing, along with aberrant subcortical automatic processing, in psychiatric disorders.

The 1990s saw the advent of functional magnetic resonance imaging (fMRI), which provided increased spatial and temporal resolution, and more within-session repeated measurements, without the use of injections or ionizing radioactivity. The Martinos Center at Massachusetts General Hospital was a pioneer in the development of this technology. Its image-acquisition and processing methods, as well as those of a few other major centers, were combined with automated, whole-brain image-analysis methods (such as Statistical Parametric Mapping, developed in London), allowing the rapid spread of cognitive neuroscience and disease-application studies to scores of major university medical centers. Work in these settings has been highly interdisciplinary, involving teams that bridge the medical, (neuro)biological, psychological, and physical sciences.

In the 2000s, the resulting investigations have created a large literature in which psychological functions, systems-level brain processes, and their disruption in psychiatric disorders are enumerated with ever increasing precision. These phenomena include elements and interactions of sensory processing, salience detection, reward processing, stress reactivity, emotional evaluation, memory, attention, decision making, goal-directed behavior, and executive control.7,8 Specific abnormalities in each (or a combination) of these have been localized and associated with subtypes of mental illness.9 Regional, modular hypo-responsivity, hyper-responsivity, biased processing, disorganized processing, dyscontrol, and compensatory activity have all been described, within and across the full range of psychiatric conditions (from personality disorders to eating disorders), in a manner that helps explain the clinical phenomenology and treatment response.10

In these studies, many of the same regions, such as amygdala and anterior cingulate cortex, have been implicated. These findings, together with the advent of multivariate connectivity analysis methods over the past few years, have informed efforts to characterize the interactions among the brain regions identified (see Figure 1). The elucidation of the resting-state default mode—and then of the task-dependent, salience, and other networks—provides a way of describing broader functional sets of distributed brain activity, along with their disruption in psychiatric disorders.11 Regionally targeted connectivity analyses are increasingly supplemented by unbiased, data-driven analyses that reveal patterns of abnormal activity underlying clinically relevant traits and states.12

Figure 1

Figure 1

These functional-neuroimaging advances have been paralleled by similar advances in MRI structural neuroimaging of psychiatric disorders. Focal abnormalities in brain volume and shape have been associated with specific types of psychotic, anxiety, and mood disorders as well as with their genetic and environmental associations. Findings include hippocampal (even subfield) atrophy, amygdalar hypertrophy, and ventromedial prefrontal neuronal versus glial volume loss. Such regional measures have been supplemented with whole-brain (voxel-based morphometry) analyses of gray matter density, cortical thickness, and structural network correlations.13 Diffusion tensor imaging techniques have permitted the assessment of the integrity of white matter tracts connecting the various brain regions into networks,14 providing a structural substrate for current connectomics research.

Such structural and functional neuroanatomic work has been accompanied by neurochemical and metabolic studies using PET radioligands with kinetic modeling and (hydrogen or phosphorus) magnetic resonance spectroscopy.15,16 The former can be specific to receptor subtypes at the sub-nanomolar level, and the latter can measure metabolic intermediates in volumes of brain tissue. This type of work has provided important insights of relevance to psychopharmacology. It is important to note that each brain-imaging method has its limitations that must be understood, its requirements for expert knowledge and interpretation, its reliance upon continued methodological development and validation (which brings new capabilities, such as 7-Tesla MRI that will allow resolution of subnuclei), its place in the translational context of animal models (which allow important basic observations but cannot provide the uniquely human information that imaging can), and its use with complementary techniques that provide greater temporal resolution (but are largely restricted to surface-based measures such as EEG/electrophysiology, magnetoencephalography, and optical imaging).17

The developments described above reflect a move in the field—from descriptive, DSM-based taxonomies, to mechanistic, neurobehavioral, intermediate-phenotype RDoC criteria, and now toward biomarker profiles, for mental illness. These imaging methods are increasingly being combined with genetic, physiological, endocrine, and other omics measures. They are being incorporated before and after therapeutic (e.g., psychotherapeutic, pharmacological, and brain stimulation) interventions. They are being used to study at-risk populations in a neurodevelopmental context. They are being deployed for longitudinal studies of larger and more diverse, stratified, well-characterized patient samples. The hope is that multimodal neuroimaging can play a critical role in the development of precision psychiatry, in which patients can be subtyped based upon biological mechanism. This subtyping would allow for the identification of risk and resilience factors, for prevention or for early detection and early (hopefully trajectory-altering) intervention, and for diagnoses that convey important prognostic information. It would also permit the identification of predictors of differential treatment response, and the development of more targeted, individualized, effective therapeutics to modulate both the identified brain circuit dysfunction and the pathophysiological processes that drive it. Indeed, just as advances in knowledge regarding neurochemistry and new molecular targets can be leveraged to develop new pharmacotherapies, advances in knowledge of psychiatric neurocircuitry provide the potential for rational development of specific neurostimulation approaches.18

It should be noted that this same panoply of tools and approaches has been employed to help us better understand the brain basis of healthy affective, sensorimotor, and cognitive functions, as well as normal human development across the lifespan. The neurocircuitry models and visualization provided by neuroimaging have also served as powerful heuristics for clinicians, patients, and families, providing a medical (and less stigmatized) framework for thinking and talking about psychiatric diseases and their treatments.

It is no hyperbole to suggest that advances of the past 25 years in neuroimaging have provided the most powerful tools to date for advancing human systems-level brain science— and psychiatry in particular. The capacity to quantitatively assay human brain structure, function, and chemistry in vivo has revolutionized our field. And yet, it is clear that progress along these lines remains in its infancy and that cost-effective clinical relevance is the ultimate goal of translation. Consequently, as exciting as the past quarter century has been, the discoveries of most profound impact—those that will change the lives of our patients and their families—still lie ahead.

Back to Top | Article Outline

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.

Back to Top | Article Outline

REFERENCES

1. Pan H, Epstein J, Silbersweig DA, et al. New and emerging imaging techniques for mapping brain circuitry. Brain Res Rev 2011;67:226–51.
2. Berman KF, Illowsky BP, Weinberger DR. Physiological dysfunction of dorsolateral prefrontal cortex in schizophrenia. IV. Further evidence for regional and behavioral specificity. Arch Gen Psychiatry 1988;45:616–22.
3. Hamani C, Mayberg H, Stone S, et al. The subcallosal cingulate gyrus in the context of major depression. Biol Psychiatry 2011;68:301–8.
4. Silbersweig DA, Stern E, Frith C, et al. A functional neuroanatomy of hallucinations in schizophrenia. Nature 1995;378:176–9.
5. Saxena S, Brody AL, Maidment KM, et al. Localized orbitofrontal and subcortical metabolic changes and predictors of response to paroxetine treatment in obsessive-compulsive disorder. Neuropsychopharmacology 1999;21:683–93.
6. Rauch SL, Whalen PJ, Shin LM, et al. Exaggerated amygdala response to masked facial stimuli in posttraumatic stress disorder: a functional MRI study. Biol Psychiatry 2000;47:769–76.
7. Posner MI, Sheese BE, Odludas Y, et al. Analyzing and shaping human attentional networks. Neural Netw 2006;19:1422–9.
8. Kensinger EA, Schacter DL. Neural processes supporting young and older adults' emotional memories. J Cogn Neurosci 2008;20:1161–73.
9. Berghorst LH, Kumar P, Greve DN, et al. Stress and reward processing in bipolar disorder: a functional magnetic resonance imaging study. Bipolar Disord 2016;18:602–11.
10. Perez DL, Vago DL, Pan H, et al. Frontolimbic neural circuit changes in emotional processing and inhibitory control associated with clinical improvement following transference-focused psychotherapy in borderline personality disorder. Psychiatry Clin Neurosci 2016;70:51–61.
11. Ambrosi E, Arciniegas DB, Madan A, et al. Insula and amygdala resting-state functional connectivity differentiate bipolar from unipolar depression. Acta Psychiatr Scand 2017;136:129–39.
12. Drysdale AT, Grosenick L, Downar J, et al. Resting-state connectivity biomarkers define neurophysiological subtypes of depression. Nat Med 2017;23:28–38.
13. Neilson E, Bois C, Gibson J, et al. Effects of environmental risks and polygenic loading for schizophrenia on cortical thickness. Schizophr Res 2017;184:128–36.
14. Saito Y, Kubicki M, Koerte I, et al. Impaired white matter connectivity between regions containing mirror neurons, and relationship to negative symptoms and social cognition, in patients with first-episode schizophrenia. Brain Imaging Behav 2017 Feb 28 [Epub ahead of print].
15. James GM, Baldinger-Melich P, Philippe C, et al. Effects of selective serotonin reuptake inhibitors on interregional relation of serotonin transporter availability in major depression. Front Hum Neurosci 2017;11:48.
16. Kim SY, Cohen BM, Chen X, et al. Redox dysregulation in schizophrenia revealed by in vivo NAD+/NADH measurement. Schizophrenia Bull 2017;43:197–204.
17. Chand GB, Dhamana M. Interactions between the anterior cingulate-insula network and the fronto-parietal network during perceptual decision-making. Neuroimage 2017;152:381–9.
18. Silbersweig DA. Depression in medical illness: future directions. In: Barksy AJ, Silbersweig DA, eds. Depression in medical illness. New York: McGraw-Hill Education, 2016:371–6.
Keywords:

brain imaging; magnetic resonance imaging; neuroimaging; positron emission tomography; psychiatric disorders

© 2017 President and Fellows of Harvard College