For this Research Topic, we invite leaders in the field of translational neuropsychiatry to submit original research and commentaries on the status of translation of fMRI into clinical practice. fMRI is now a ubiquitous technology available at imaging centers small and large around the world, and is used in studies of myriad brain and mental processes. Yet, after 25 years of methods development and application, a clear path to incorporating this neuroimaging technique in the diagnosis, prognosis, or treatment of mental illness remains elusive. We will explore, from various perspectives, the current barriers to translation and how they may be addressed, across several possible bases:
• neurobiological
• computational
• biobehavioral
• neurodevelopmental
• physical
• pharmacological
• nosological
• ontological
• practical; and
• socioeconomic.
These factors will be addressed towards creating a roadmap to realms of translational application for fMRI and the limits of its application.
Numerous neuroimaging and electrophysiological imaging techniques are currently available to study neural function at various levels, from synaptic to circuit, man and animal, invasiveness and so on. Despite this plethora of techniques, many of exquisite sensitivity (e.g. calcium imaging), fMRI is still the only technique that can noninvasively image the entire animal or human brain via a contrast mechanism relatively close to neuronal activity at spatial and temporal scales suitable for regional, genomic, circuit-level investigation. fMRI is widely available, applicable across a wide range of conditions, and relatively low cost. To date, there is no viable alternative that can, point-for-point, surpass fMRI in its potential for translational value. Even when using reference data from invasive and electrophysiological measures, repeated studies over the years show that neurovascular hemodynamics in response to neuronal demand are a powerful and reliable correlate of neurophysiology as well as of bio-behavior. Indeed, fMRI - resting-state or task-based - at its best, can offer a wealth of information on whole-brain function in an extremely wide range of applications.
To what then can we attribute the failure to find meaningful markers of neuropsychiatric disease and how can we remedy these limitations? We will address issues related first to measurement of physiological suitability of clinical fMRI to neuropsychiatric disease; encoding of diagnostic and prognostic information in fMRI data, and those related. We will, as an example, evaluate the implications of current dependence on linear systems to describe neural processes correlates of disease, and to what extent this is appropriate or limiting for assessment of markers of neuropsychiatric disease. The state of nonlinear or latent or causal processes will be investigated as a basis of decoding brain-behavior relationships in neuropsychiatric disease. In a different vein, nosology and ontology will be addressed either as an instrument or a hindrance in addressing the relationship of disease to brain activity, and whether a tabula rasa data-driven approach may be enabling in this pursuit. In the same manner, statistical limitations including sampling and N=1 studies for the discovery and application of markers will be explored.
The regular application of neuroimaging in the clinical evaluation and treatment of neuropsychiatric disease seems ever closer to reality, yet much work needs to be done to realize this vision. Given the successful culmination of communicating the collection of articles envisioned in this Research Topic, readers and the field more generally will have a better grasp on next steps in the transformational integration of neuroimaging in psychiatric practice.
For this Research Topic, we invite leaders in the field of translational neuropsychiatry to submit original research and commentaries on the status of translation of fMRI into clinical practice. fMRI is now a ubiquitous technology available at imaging centers small and large around the world, and is used in studies of myriad brain and mental processes. Yet, after 25 years of methods development and application, a clear path to incorporating this neuroimaging technique in the diagnosis, prognosis, or treatment of mental illness remains elusive. We will explore, from various perspectives, the current barriers to translation and how they may be addressed, across several possible bases:
• neurobiological
• computational
• biobehavioral
• neurodevelopmental
• physical
• pharmacological
• nosological
• ontological
• practical; and
• socioeconomic.
These factors will be addressed towards creating a roadmap to realms of translational application for fMRI and the limits of its application.
Numerous neuroimaging and electrophysiological imaging techniques are currently available to study neural function at various levels, from synaptic to circuit, man and animal, invasiveness and so on. Despite this plethora of techniques, many of exquisite sensitivity (e.g. calcium imaging), fMRI is still the only technique that can noninvasively image the entire animal or human brain via a contrast mechanism relatively close to neuronal activity at spatial and temporal scales suitable for regional, genomic, circuit-level investigation. fMRI is widely available, applicable across a wide range of conditions, and relatively low cost. To date, there is no viable alternative that can, point-for-point, surpass fMRI in its potential for translational value. Even when using reference data from invasive and electrophysiological measures, repeated studies over the years show that neurovascular hemodynamics in response to neuronal demand are a powerful and reliable correlate of neurophysiology as well as of bio-behavior. Indeed, fMRI - resting-state or task-based - at its best, can offer a wealth of information on whole-brain function in an extremely wide range of applications.
To what then can we attribute the failure to find meaningful markers of neuropsychiatric disease and how can we remedy these limitations? We will address issues related first to measurement of physiological suitability of clinical fMRI to neuropsychiatric disease; encoding of diagnostic and prognostic information in fMRI data, and those related. We will, as an example, evaluate the implications of current dependence on linear systems to describe neural processes correlates of disease, and to what extent this is appropriate or limiting for assessment of markers of neuropsychiatric disease. The state of nonlinear or latent or causal processes will be investigated as a basis of decoding brain-behavior relationships in neuropsychiatric disease. In a different vein, nosology and ontology will be addressed either as an instrument or a hindrance in addressing the relationship of disease to brain activity, and whether a tabula rasa data-driven approach may be enabling in this pursuit. In the same manner, statistical limitations including sampling and N=1 studies for the discovery and application of markers will be explored.
The regular application of neuroimaging in the clinical evaluation and treatment of neuropsychiatric disease seems ever closer to reality, yet much work needs to be done to realize this vision. Given the successful culmination of communicating the collection of articles envisioned in this Research Topic, readers and the field more generally will have a better grasp on next steps in the transformational integration of neuroimaging in psychiatric practice.
