We searched PubMed from January, 1990, to August, 2013. Search items were combinations of “stroke”, “neuroimaging”, “functional connectivity”, “effective connectivity”, “DTI”, and “motor system”. For some sections of this Review, additional keywords such as “animal”, “rat”, “monkey”, and “transcranial magnetic stimulation”, “electroencephalography”, or “magnetoencephalography” were used. The final reference list is based on the relevance of the articles to the scope of this Review.
ReviewConnectivity-based approaches in stroke and recovery of function
Introduction
Stroke is a leading cause of long-term disability (WHO atlas of heart disease and stroke 2004).1 In countries with well developed health care systems, stroke-associated mortality has continuously declined in the past decade because of improvements in acute stroke treatment (eg, recanalisation therapy, decompressive therapy) and medical care (stroke units).1 The increasing proportion of stroke survivors is, however, associated with a growing number of patients living with a persistent neurological deficit; despite intensive rehabilitation, more than half of all stroke patients are greatly disabled.1 Many of these patients show persistent motor symptoms, which affect their functional independence in everyday life.2 In view of our ageing societies, the burden of stroke is expected to rise further in the next decades, thus an urgent need emerges to further our understanding of the neurobiological factors that determine functional outcome to inform novel treatment approaches. Functional neuroimaging paves the way for non-invasive insights into the neural mechanisms underlying recovery of function and reorganisation of brain networks.
Importantly, recent developments in computational neuroscience enable us to move beyond the mere localisation of brain activity. In particular, they allow us to consider the dynamics within an ensemble or an entire network of areas sustaining a particular cognitive process or behaviour. Such analyses open up new vistas on the pathophysiology underlying stroke-induced neurological deficits and the network changes underlying recovery of function.
In this Review, we discuss recent data obtained by neuroimaging experiments that provide new insights into the mechanisms underlying recovery of function from a systems-level approach. We first summarise data obtained from animal studies that show physiological mechanisms engaged in functional recovery. We next review novel methods that non-invasively assess connectivity of brain areas and changes thereof during cortical reorganisation in patients who have had a stroke. We focus on MRI-based imaging techniques, which because of their excellent spatial resolution, enable us to study the contribution of distinct anatomical areas to recovery. Other effective brain mapping methods such as electroencephalography, magnetoencephalography, and transcranial magnetic stimulation are also briefly discussed; a more elaborate assessment of these electrophysiological methods is however beyond the scope of this Review. Finally, we aim to reconcile the findings obtained by different approaches (including functional MRI [fMRI], electrophysiological studies, and animal data) to provide a comprehensive picture of reorganisation processes after stroke. Such neuroimaging-based analyses could eventually be used to inform novel treatment strategies for neurorehabilitation.
Section snippets
Stroke and disconnection concepts in the nervous system
Nearly 100 years ago, the Russian-Swiss neurologist Constantin von Monakow coined the concept of diaschisis, which postulates that an acute lesion to a part of the brain consecutively leads to a reduction of excitatory input into regions remote from but connected to the lesion.3, 4 The resulting depression of the functionality of interconnected regions (so-called passive inhibition) was assumed to contribute to the neurological deficit of the patient.5 Von Monakow further hypothesised that
Spontaneous recovery
Animal studies have shown many biochemical and cellular processes triggered by stroke. For example, inflammatory responses such as activation of glial cells, cytokines, and other immunomodulators, and activation of neural stem cells and changes in genetic machinery, lead to enhanced expression of neuroprotective proteins, nerve growth factors, and neurotransmitter receptors.8, 9, 10, 11 Importantly, these effects take place within minutes and hours after a stroke, and do not only occur in the
Neuroimaging of activity and connectivity
The development of non-invasive functional imaging techniques has greatly advanced our understanding of the neural mechanisms underlying behaviour and its disturbances after brain lesions in humans. These techniques not only enable us to directly test in vivo the relation between structural or functional disruptions and clinical manifestations of disease, but also warrant the opportunity for multiple testing and monitoring of treatment effects. In particular, PET and fMRI are frequently used to
Imaging motor recovery
Both PET and fMRI have been used to investigate changes in neural networks after brain lesions. A frequent finding is that ischaemic lesions alter neural activity in both the affected and the unaffected hemisphere (figure 1). Studies of animal models and patients who have had a stroke show that in the first few days after stroke, brain activity is typically reduced in the lesioned hemisphere.39, 40, 41 Thereafter, neural activity gradually increases concurrent to functional recovery, both in
Neuroimaging of structural connectivity
A key factor affecting brain networks after stroke is the anatomical damage. Lesion location rather than the mere size of the lesion accounts for the neurological sequelae after stroke.56, 57 For example, severe hemiplegia could be caused by a small lesion confined to the posterior limb of the internal capsule. This anatomically specific effect arises from disruption of the corticospinal tract fibres connecting cortical motor areas with motor neurons in the spinal cord.58 Diffusion-tensor
Neuroimaging of functional connectivity
The widespread but specific structural changes observed in patients who have had a stroke are mirrored by distinct changes of functional interactions between cortical areas. In neuroimaging, functional connectivity between brain regions can be measured in two different functional states: during a particular task or in the absence of a structured task—ie, during rest.31 During rest, participants are requested to lie motionless in the scanner without thinking of something particular (but to stay
Functional connectivity and recovery
Functional connectivity analyses based on resting-state fMRI have identified stroke-induced disturbances of the functional network architecture in both animals and patients (figure 2B). For example, resting-state measurements in rats recovering from induced stroke showed that impaired sensorimotor performance was associated with a loss of interhemispheric connectivity between sensorimotor regions, whereas recovery of function weeks after stroke was paralleled by normalisation of
Effective connectivity after stroke
A severe restriction of functional connectivity analyses is that they do not provide information about the directionality of functional interactions. By contrast, models of effective connectivity explicitly test the effect that one area exerts on another.28 So far, most studies on effective connectivity changes after stroke focus on dynamic causal modelling of fMRI activity.
Dynamic causal modelling applied to fMRI data obtained from healthy individuals suggests that movements of the right or
Effective connectivity and recovery
Longitudinal studies of changes of fMRI effective connectivity in stroke showed that in the first few days after ischaemia, coupling of the ipsilesional supplementary motor area and ventral premotor cortex with ipsilesional M1 was significantly reduced.82 Coupling parameters between these areas increased with recovery and predicted a better outcome 3–6 months later. Thus, there seems to be a tight relation between changes of motor system activity, premotor–M1 connectivity, and early recovery
Treatment implications
The question of whether or not contralesional areas support recovery of function is highly relevant with respect to the development of new treatment approaches. For example, non-invasive brain stimulation techniques such as repetitive transcranial magnetic stimulation or transcranial direct current stimulation can be used to either enhance or suppress neural activity of the stimulated region.50, 52, 90, 91 However, so far, data are inconsistent about the treatment efficacy of these approaches;
Conclusions and future directions
Connectivity-based analyses of neuroimaging data allow new insights into the pathophysiology underlying stroke-induced deficits, as they provide an in-vivo systems-level perspective of the specific outcomes that a lesion has on neural networks. Therefore, these approaches do not only enable us to test established (but mostly theoretically founded) concepts of disconnection syndromes in an experimental setting, but also to extend these by providing the opportunity to study recovery of function
Search strategy and selection criteria
References (97)
Constantin von Monakow (1853–1930): a pioneer in interdisciplinary brain research and a humanist
C R Biol
(2006)- et al.
Morphology of somatosensory evoked fields: inter-hemispheric similarity as a parameter for physiological and pathological neural connectivity
Neurosci Lett
(2000) - et al.
State-dependent differences between functional and effective connectivity of the human cortical motor system
Neuroimage
(2013) - et al.
Dynamic causal modelling
Neuroimage
(2003) - et al.
Diffusion MRI at 25: exploring brain tissue structure and function
Neuroimage
(2012) - et al.
A longitudinal fMRI study: in recovering and then in clinically stable sub-cortical stroke patients
Neuroimage
(2004) - et al.
Activation likelihood estimation meta-analysis of motor-related neural activity after stroke
Neuroimage
(2012) - et al.
Exploring Theta Burst Stimulation as an intervention to improve motor recovery in chronic stroke
Clin Neurophysiol
(2007) - et al.
The behavioral significance of coherent resting-state oscillations after stroke
Neuroimage
(2012) - et al.
Resting state α-band functional connectivity and recovery after stroke
Exp Neurol
(2012)