Functional and structural reorganization in relation to functional outcomes after stroke : insights from magnetic resonance imaging

Abstract

In view of the great impact of stroke, the need and importance of good recovery has been highlighted. Currently, based on the evidences stating the potential for neuroplasticity and its relation to stroke recovery, a restorative and neuroscience-based approach to stroke rehabilitation, such as motor imagery training, has been developed and recommended to the rehabilitation specialists. However, the mechanism of the shaping of neuroplasticity in relation to stroke recovery is still not fully understood. In addition, a number of factors, such as the extent of damage in influencing the structural and functional brain networks, have been identified to have certain impact on the potential for neuroplasticity and the capacity for receiving benefits from rehabilitation. Thus, more understanding of the neural mechanism that underpins an effective and restorative approach to stroke rehabilitation becomes necessary and important. The emerging of neuroimaging techniques can therefore help to study this mechanism of post-stroke recovery. Magnetic resonance imaging (MRI), which is one of these techniques, has been shown having a great potential to give important insights into post-stroke recovery mechanisms by providing rich information on both the structural and functional aspects of the brain. The overall objective of this study is to find out how the functional outcomes after stroke be related to the changes in functional networks responsible for motor execution (ME) and motor imagery (MI), and the changes in physical structure of the brain. This study is split into two parts to look into the features of brain reorganization and remodeling in patients after stroke from both functional and structural perspectives, and to correlate these features with their functional outcomes. Part I of this study is going to explore neural correlates of motor impairment during ME and MI from a functional perspective using functional magnetic resonance imaging (fMRI), while part II of this study is going to describe remodeling of structural connectivity and its correlation with motor impairment from a structural perspective using diffusion tensor imaging (DTI). In study I, ten chronic stroke patients with left subcortical ischemic lesions and right hemiparetic limbs, and ten unimpaired subjects were included. Their cortical processes were studied when they were asked to perform ME and MI unimanually using their unaffected and affected wrists during fMRI. Laterality index (LI) and overlap index (OI) were used to quantify hemispheric asymmetry and the spatial discrepancy of the cerebral activations, respectively. In study II, one of the stroke patients was excluded, since the anatomical image of this patient was contaminated by motion artifacts, which led to failure in subsequent data analysis. One of the unimpaired subjects with similar age was also excluded to match the total number of stroke patients included in this study. Therefore, only nine chronic stroke patients having left subcortical ischemic lesions and nine unimpaired subjects were included to study the post-stroke structural remodeling based on the data collected using DTI. Three connectivity measures (fractional anisotropy FA, connection weight CW, connection strength CS) were used to localize the fiber tracts and areas that were affected by the lesions. From correlation results in study I, the supplementary motor area (SMA), its activation volume and congruence in functional neuroanatomy associated with ME and MI using affected wrist positively correlated with motor performance. During ME of affected wrist, the precuneus, its activation volume and congruence in functional neuroanatomy between patient and unimpaired groups showed negative correlation, while in non-primary motor areas, the hemispheric balance of premotor cortex and the congruence in functional neuroanatomy of contralesional inferior parietal lobule between patient and unimpaired groups showed positive correlation with motor performance. From the results in study II, significant positive CS-motor correlations were found in five common regions (ipsilesional precentral gyrus, ipsilesional rostral middle frontal cortex, ipsilesional pallidum, ipsilesional amygdala and contralesional lingual gyrus) before and after controlling for the bilateral CSTs' connectivity property. Meanwhile, the CSs of three of these regions (precentral gyrus, rostral middle frontal cortex and amygdala) were also found significantly higher in unimpaired subjects compared with those in stroke patients, indicating that these three regions exhibited weaker connections to the rest of the network for the stroke patients. From a functional perspective, the non-primary motor-related areas were revealed to play a critical role in determining motor outcomes after left subcortical stroke, which was demonstrated in our stroke patients. In particular, SMA might be the key neural substrate recruited during MI, which is in association with motor recovery. From a structural perspective, statistically significant alterations of neural connectivity in both the ipsilesional and contralesional hemispheres were shown in our stroke patients compared to the unimpaired subjects, implying structural remodeling occurred in widespread areas in both the ipsilesional and contralesional hemispheres.