ReviewImaging human brain networks to improve the clinical efficacy of non-invasive brain stimulation
Introduction
Healthy brain function relies on co-ordinated integration of localized activity across widespread neural networks (Catani et al., 2013, Park and Friston, 2013). Understanding how such activity is integrated globally across the brain is arguably one of the greatest challenges facing modern neuroscience (Devor et al., 2013). Although the function and physiological impact of neural activity in specialized brain regions is relatively well understood, we are only beginning to understand how this local activity is integrated between distant regions – for example between primary sensory and higher cognitive areas – or how local damage to, or stimulation of, a specific neural population affects activity elsewhere in the brain.
Our understanding of how local changes in brain activity can influence distant, but functionally related, brain regions has improved in parallel with advances in various forms of brain imaging and brain stimulation methods. For example, stroke patients with local brain lesions often have cognitive impairments that cannot be directly related to the site of damage (Verdon et al., 2010). Further, many of the most common psychiatric and neurological conditions, including depression, obsessive-compulsive disorder and schizophrenia, are associated with impaired integration of functionally related neural networks (Insel, 2010, Menon, 2011, Zhou et al., 2012, Filippi et al., 2013, Fornito et al., 2015).
In recent years it has become clear that various forms of non-invasive brain stimulation (NIBS), such as transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (tES), can modify ongoing brain activity. This has led to a dramatic increase in research applying NIBS in the clinical domain, with the goal of improving abnormal brain function in various conditions (Hummel et al., 2005, Hummel and Cohen, 2006, Passard et al., 2007, Floel, 2014, Liew et al., 2014). The rationale for the use of NIBS has been that if behavioral changes arising from a clinical condition occur due to altered activity within a given brain network, normalizing this activity with NIBS should lead to improved behavior. Such a rationale has motivated studies utilizing NIBS across a range of clinical conditions, including, but not limited to, stroke (Grefkes and Fink, 2014), schizophrenia (Frantseva et al., 2014), depression (Fitzgerald et al., 2009, Fox et al., 2013), and obsessive–compulsive disorder (Fitzgerald et al., 2010). Despite some encouraging results, symptomatic improvement following NIBS in such conditions has generally been modest, and often clinically insignificant (Kalu et al., 2012). Although several reviews have alluded to the effects of NIBS propagating to distant regions via axonal connectivity (Lefaucheur et al., 2014, Liew et al., 2014, Li et al., 2015), they have not considered the consequence of these aspects in the clinical domain. An often overlooked factor affecting the effectiveness of NIBS is the optimal targeting of underlying neural networks associated with the clinical condition (Fox et al., 2012). For instance, the efficacy of TMS in the treatment of depression is influenced by the precise location of stimulation within the DLPFC. The region with the highest efficacy was a region with anticorrelated activity between DLPFC and the subgenual cingulate (Fox et al., 2012). Further, several studies have now investigated how each of the different forms of NIBS alters the activity of large-scale neural networks. Studies suggest that the global effect induced by NIBS is dependent on several factors including electrode placement (Sehm et al., 2013), connectedness of the targeted brain region to other regions (e.g., Rounis et al., 2006 vs. Cocchi et al., 2015), and whether the participant is concurrently undertaking a task (Nitsche et al., 2003, Fregni et al., 2005, Hummel et al., 2005, Antal et al., 2007, Kujirai et al., 2006, Andrews et al., 2011). This interplay between functional interactions between brain regions and NIBS in health and disease has been the focus of many recent studies.
In this review, we discuss prevailing ideas regarding brain function as a complex interaction between multiple neural networks, and outline how different forms of NIBS might interact and modify these networks. We then relate this to how pathological network disturbances could be optimally targeted with specific forms of NIBS. When applying NIBS in the clinical domain, we contend that without a thorough understanding of both the specific network disturbance of the targeted condition, and the mode of action of the NIBS protocol, the desired benefits may not be obtained, or worse, any effects on functioning may be detrimental.
Section snippets
Local specialization versus global integration of brain function
In the last few years, significant funding has been invested in research attempting to characterize the structure and function of the human brain as a network (Hagmann et al., 2008, Betzel et al., 2014, Sporns, 2014a, Sporns, 2014b). Using neuroimaging – particularly functional magnetic resonance imaging (fMRI) – and metrics borrowed from the mathematical field of graph analysis, connectomics has started to reveal the complex principles that support flexible brain function (Bullmore and Sporns,
Brain stimulation as a clinical intervention for restoring network function
Several stimulation techniques can induce changes in local neural activity, and can therefore be potentially useful in modulating dysfunctional neural networks. These stimulation techniques can be broadly divided into two main forms – TMS and electrical brain stimulation. Electrical brain stimulation includes deep brain stimulation (DBS), electroconvulsive therapy, and tES. Because of their relatively non-invasive nature, this review will focus on TMS and tES as clinical interventions for
Connectomics as a tool to guide the clinical use of NIBS
To enhance its clinical efficacy, it will be important to establish which conditions and symptoms respond favorably to NIBS, which brain regions should be targeted to maximize efficacy, and which NIBS intervention might be most effective to do so. Intuitively, a diffuse pathology affecting large cortical territories or widespread cortical networks might benefit most from a non-focal NIBS approach, whereas a focal lesion might respond most effectively to a locally mediating NIBS intervention
Context dependent modulation of neural networks
In the majority of published studies on the effects of rTMS and tDCS on brain activity, participants were not required to perform a behavioral task (i.e., stimulation was delivered at rest; offline). The idea behind this approach is to minimize variability in the excitability of the stimulated neurons by keeping participants in a uniform state of rest, as changes in the activity/excitability of neurons targeted by NIBS can influence how those neurons respond to stimulation (Silvanto et al., 2007
Modulating local cortical synchronization
The task-related boosting (or suppression) of neural oscillations at specific frequencies between cortical regions can enhance performance in processes involved in task selection, attention and memory (Marshall et al., 2004, Lakatos et al., 2008, Bauer et al., 2012, Helfrich et al., 2014). Both rTMS and tDCS can modify ongoing oscillatory activity, and in so doing, may mimic or augment the amplitude of normal endogenous neural oscillations to promote task performance and cortical connectivity.
Implications for the treatment of brain disorders
NIBS has been shown to induce significant changes in functional connectivity and performance on a variety of functional tasks. As noted above, there are important differences in how the different stimulation paradigms modify brain function. Such differences suggest that NIBS methods can be selected to target known alterations in large-scale brain network activity in a variety of psychiatric and neurological conditions. Combining knowledge on the mechanisms of each NIBS method with emerging
Other considerations
We have focused on the relative merits of rTMS, tDCS and tACS for restoring functional networks affected by a range of psychiatric and neurological conditions. For any particular stimulation approach, identifying where to stimulate (e.g., hub vs. non-hub regions of cortex) is likely to be critical (Fox et al., 2014). Given the specific nature of the network disturbance in various brain disorders, one form of NIBS may be more efficacious than others. When considering the use of brain stimulation
Final remarks
The treatment of most psychiatric and neurological conditions is complex, expensive, and often requires multimodal interventional strategies. Based on the evidence reviewed here, we argue that non-invasive brain stimulation techniques represent an important, complementary approach for the treatment of psychiatric and neurological disorders. Behavioral therapies are efficient but slow, and often require a reasonable level of insight and co-operation from patients. Thus, for severe psychiatric
Acknowledgements
MVS was supported by an NHMRC Australia-based Biomedical Postdoctoral Research Fellowship (APP1012153) and NHMRC Project Grant (GNT1078464). JBM was supported by the ARC-SRI Science of Learning Research Center (SR120300015), the ARC Center of Excellence for Integrative Brain Function (ARC Center Grant CE140100007), and an ARC Australian Laureate Fellowship (FL110100103). AZ was supported by an NHMRC Career Development Fellowship (GNT1047648). The authors wish to thank Dr. David Lloyd with his
References (167)
- et al.
The effects of high-frequency rTMS over the left dorsolateral prefrontal cortex on reward responsiveness
Brain Stimul.
(2013) - et al.
Improving working memory: the effect of combining cognitive activity and anodal transcranial direct current stimulation to the left dorsolateral prefrontal cortex
Brain Stimul.
(2011) - et al.
Comparatively weak after-effects of transcranial alternating current stimulation (tACS) on cortical excitability in humans
Brain Stimul.
(2008) - et al.
Brain-derived neurotrophic factor (BDNF) gene polymorphisms shape cortical plasticity in humans
Brain Stimul.
(2010) - et al.
Left DLPFC rTMS stimulation reduced the anxiety bias effect or how to restore the positive memory processing in high-anxiety subjects
Psychiatry Res.
(2013) - et al.
Cholinergic enhancement of visual attention and neural oscillations in the human brain
Curr. Biol.
(2012) - et al.
Repetitive transcranial magnetic stimulation (rTMS) for obsessive–compulsive disorder (OCD): an exploratory meta-analysis of randomized and sham-controlled trials
J. Psychiatr. Res.
(2013) - et al.
Evidence for motor cortex dedifferentiation in older adults
Neurobiol. Aging
(2012) - et al.
Changes in structural and functional connectivity among resting-state networks across the human lifespan
Neuroimage
(2014) - et al.
Connectomics reveals faulty wiring patterns for depressed brain
Biol. Psychiatry
(2014)