Synopsis

Keywords: Brain Connectivity, fMRI (resting state)

High-resolution resting-state (rs) fMRI enables the functional mapping of vein-dominated connectivity, correlated with neuronal calcium signals. Here, we performed high-resolution rs-fMRI to examine 30 ischemic stroke patients with binocular isotropic hemianopia and 10 healthy controls. The results of seed-based and independent component analysis (ICA) demonstrated the reduced vein-dominated correlation patterns. Interestingly, ICA also found increased correlations between cerebrospinal fluid (CSF) and ischemic lesions in stroke patients, showing ultra-slow oscillation frequencies up to 0.04 Hz. Our findings suggest that ischemic lesions are associated with CSF, prompting disruption of vein-dominated connectivity.

Introduction

Stroke is a fatal disease with high morbidity, disability and recurrence rates, closely related to a focal or widespread loss of blood flow [1]. Our previous work performed high-resolution single-vessel fMRI to map spatiotemporal correlations of venous BOLD fMRI fluctuations, demonstrating the vein-dominated connectivity correlated with neuronal intracellular calcium signals [2]. Here, we used high-resolution resting-state functional magnetic resonance imaging (fMRI) to investigate how ischemic lesions affect vessel-specific connectivity in the human stroke brain.

Methods

All MR images were acquired on a 3T GE MRI scanner with a 32-channel head coil. We examined 30 ischemic stroke patients with homonymous hemianopia (male/female: n=23/7; mean±SD: 60.5 ± 8.8 years) and 9 healthy controls (male/female: n=3/6; mean±SD: 64.4 ± 6.8 years).
High-resolution Resting-state fMRI images: we obtained resting-state fMRI images covering the occipital lobe in the coronal orientation with the following parameters: TR=4500 ms, TE=30 ms, FA=90°, matrix = 256 × 256, FOV = 256 ×256 mm², in-plane resolution = 1×1 mm² and 21 slices with thicknesses of 1.5 mm.
Diffusion MRI (dMRI) images: a diffusion-weight EPI sequence was employed in the same range with fMRI using b=1000 and 2000 s/mm², TR = 5600 ms, TE = 100 ms, matrix = 128×128, FOV = 256 × 256 mm², in-plane resolution = 2 x 2 mm², and 16 slices with thickness = 4.5 mm. All data were processed and analyzed by AFNI, MRtrix3, and Independent Component Analysis (ICA) Toolbox (GIFT 4.0).

Results

Disrupted vein-dominated connectivity in stroke patients. As our previous work, the seed-based correlation maps of healthy controls demonstrated vein-dominated correlation spatial patterns (Fig. 1 B-E). However, when we selected individual veins around the infarct as seeds, the seed-based analysis exhibited weak connectivity with other veins (Fig. 1G-H). Furthermore, the seed-based correlation maps show similar venule-specific correlation patterns only in the hemisphere without lesions (Fig. 1I-J).
To avoid selection bias in seed-based analysis, we performed an ICA analysis in stroke patients. The first component (Com1) maps of ICA demonstrated vein-dominated patterns in the normal hemisphere while dysconnectivity in the hemisphere with lesions, which are consistent with the seed-based analysis in stroke patients (Fig. 2). The 3D reconstruction clearly showed the decreased vessel-specific connectivity induced by infarcts in stroke patients. (Fig. 3A). The mean power spectrum density (PSD) of the ICA-Com1 (n=30) indicated the ultra-slow oscillation up to 0.04 Hz (Fig. 3B-C). The venous component ratios in ROI-Infarct are significantly lower than that in ROI-normal (*, p<0.05, n = 10, Fig. 3D-E).
The increased functional connectivity between CSF and ischemic lesions in stroke patients. We also found one interesting ICA component in stroke patients, showing the correlation patterns between ischemic lesions and CSF (Fig. 4). The 3D ICA reconstruction showed CSF-dominated correlation patterns (ICA-Com2) overlaid with lesions (Fig. 5A). BOLD fMRI from the CSF and the lesions fluctuated at the ultra-slow frequency to 0.04 Hz (Fig. 5B-C). Furthermore, the CSF component ratios in ROI-Infarct are significantly higher than that in ROI-normal (*, p < 0.05, n = 10, Fig. 5D-E).

Discussion

Mestre et al. [3] implemented multimodal in vivo imaging in rodents and demonstrated that CSF rushes into the rodents' brains, causing swelling after a stroke. There is still no in-vivo human study to support it. Here, we used high-resolution rs-fMRI to further confirm that ischemic lesions are associated with CSF in the human brain, prompting disruption of vein-dominated connectivity.

Conclusion

We performed high-resolution rs-fMRI to reveal brain connectivity dysfunction in stroke patients. We found reduced vessel-specific connectivity induced by infarcts but increased functional connectivity between CSF and ischemic lesions in stroke patients. CSF and lesions shared similar ultra-slow oscillation up to 0.04 Hz, suggesting that ischemic lesions are associated with CSF, prompting disruption of vein-dominated connectivity. The study demonstrated the feasibility of high-resolution rs-fMRI in the functional assessment of stroke, opening the way for vessel-specific connectivity in stroke assessment.

Acknowledgements

We thank the Analysis of Functional NeuroImages (AFNI) team for software support. This work was supported by grants from the National Natural Science Foundation of China (No. 82201447), the Department of Science and Technology of Guangdong Province (2018B030322006), the Hundred Talents Program of Sun Yat-sen University (The Fifth Affiliated Hospital, 202101).