Introduction

Resting-state functional connectivity approach has been widely used in stroke patients which could provide insights into the neurological mechanisms. However, lacking the understanding of physiological mechanisms limits its interpretation of relevant findings. Artery spin labeling (ASL) can not only noninvasively measure regional cerebral blood flow (rCBF) , but is also well-coupled with brain metabolism 1. Combing functional connectivity (FC) with rCBF can reveal neurological and metabolism mechanism as well. Previous studies has investigated the relationship between rCBF and FC in healthy controls and schizophrenia patients2,3, but to our knowledge, no one has done this in stroke patients before. This study investigates the relationship between rCBF derived from ASL and FC derived from Blood oxygen level dependent (BOLD) at the recovery phase in 21 stroke patients with unilateral middle cerebral artery (MCA) infarction. Partial correlation has carried out with age and gender as covariates.

Methods

Written informed consent was obtained from all patients and all protocols were approved by the local Institutional Review Board. 21 stroke patients (age=63.19±11.24, F/M=6/15) were recruited in this project. All patients underwent Modified Rankin Scale (mRS) measurements three months after stroke first onset on the same day of MRI scan. Except four patients with aphasia who did not cooperate to accomplish the cognition assessment, the others all completed Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA). 3D T1, artery spin labeling and blood oxygen level dependent imaging were carried out on a 3T Siemens Trio Tim scanner using 32-channle head coil. Imaging analysis was performed using Data Processing Assistant for Resting-State fMRI (DPARSFA), Statistical Parametric Mapping (SPM), graph-theoretical network analysis (GRETNA) and REST. T1 images were coregistered to BOLD images followed by ASL coregistered to coregistered T1 images, making it possible to stay in consistent. Cost–function modification method and individual mask was used to adjust for the distortion of lesioned tissue4. CBF was obtained through ASL image, while FC was calculated based on the temporal similarities between BOLD signals by applying a threshold to estimate if a connection existed. AAL template was applied to divide the grey matter into 45 regions in each hemisphere. The correlation among rCBF, FC and clinical measurements were calculated between both hemispheres of grey matter using partial correlation with age and gender as covariates.

Results

Whole brain CBF is positively correlated with FC in whole brain, inter-hemispheric and intra-hemispheric (r=0.61, 0.63, 0.57 respectively). Besides, whole brain CBF is positively correlated with MMSE (r=-0.61, p=0.017), and was negatively correlated with mRS (r=0.55, p=0.035). In ipsilateral hemisphere (lesion side), the mean CBF was positively correlated with the FC strength of the whole brain (r=0.63, p=0.004) and that of ipsilateral hemisphere (r=0.73, p<0.001). Furthermore, the mean CBF of ipsilateral hemisphere was positively correlated with MMSE (r=60, p=0.019) and MoCA (r=0.62, p=0.014), but negatively correlated with mRS (r=-0.70, p=0.004). No significant correlation was found among CBF, FC and clinical scores in contralateral hemisphere (healthy side).

Discussion

Our findings demonstrated a close relationship between rCBF and FC strength in stroke patients with unilateral middle cerebral artery infarction. The cerebral circulation supplies the brain with nutrients, and is meticulously adjusted to meet the metabolic demands of the tissue. Previous studies has found that CBF coupled well with brain metabolism like aerobic glycolysis and oxygen consumption [3]. Deoxyhemoglobin decreases in the brain during activation, resulting in an increased MR signal, which can be detected by blood oxygen level dependent imaging5. The altered metabolism caused by brain activity can be measured by both ASL and BOLD, which would explain the consistence of altered tendency of CBF and FCs in ipsilateral hemisphere. Furthermore, we have found that there is a correlation between clinical assessment and CBF, indicating better prognosis is correlated with more cerebral blood flow.

Conclusion

The mean CBF in ipsilateral hemisphere but not in the unaffected hemisphere is correlated with the function of brain network and the cognitive function during recovery phase in patients with unilateral MCA infarction. These findings provide important evidence about improving the CBF of the affected hemisphere to help the functional recovery in stroke patients.

Acknowledgements

No acknowledgement found.