Visualization of perivascular spaces in the human brain at 7 T: sequence optimization and morphology characterization
Introduction
Perivascular spaces (PVSs), also known as the Virchow–Robin spaces, are fluid-filled spaces surrounding penetrating arteries and veins in the brain (Esiri and Gay, 1990). While enlarged PVSs are commonly observed in MR images in a number of neurological disorders (Conforti et al., 2013, Doubal et al., 2010, Inglese et al., 2005, Li et al., 2014, Potter et al., 2013, Rouhl et al., 2012, Wardlaw, 2010), normal PVSs are typically not visible due to their small sizes, particularly in young adults; histopathological findings have suggested a positive relation between the dimensions of PVS and age (Pesce and Carli, 1988). However, the physiological and pathophysiological significances of PVS remain elusive. Recently, several lines of evidence have suggested that PVS is a part of the brain “lymphatic” system through which interstitial solutes are cleared from the brain (Iliff et al., 2013, Kress et al., 2014, Rangroo Thrane et al., 2013, Yang et al., 2013). Specifically, it has been demonstrated that arterial pulsation drives subarachnoid cerebral spinal fluid (CSF) flow into the PVS, clearing soluble proteins such as amyloid beta (Aβ) from the brain (Bilston et al., 2003, Iliff et al., 2013). Dysfunction of PVS pathways thus may lead to enlarged PVS, increased Aβ deposition, and subsequent neuronal dysfunction and loss, which clearly have profound implications in Alzheimer's disease (Iliff et al., 2012, Yang et al., 2013). While these recent studies have provided invaluable insights into the functions of PVS for cleaning interstitial solutes, they employed invasive approaches such as two-photon microscopy or infusion of fluorescent and radio-labeled tracers that are not applicable to humans (Iliff et al., 2013, Kress et al., 2014, Rangroo Thrane et al., 2013, Yang et al., 2013).
Noninvasive imaging of PVS morphology may provide important insights into the functional status of PVS. Both T1- and T2-weighted MRI sequences have been employed for imaging PVS (Bouvy et al., 2014, Maclullich et al., 2004, Wuerfel et al., 2008, Zhu et al., 2011). However, previously reported results have largely focused on abnormal (“dilated” or “enlarged”) PVS using relatively low-resolution images with respect to the diameters of normal PVSs which are typically in the range of 0.13–0.96 mm, with the majority below 0.5 mm (Pesce and Carli, 1988). As a result, the normal morphological features of PVS remain elusive. The increased signal-to-noise ratio (SNR) at 7 T enables acquisition of high resolution anatomical images and could potentially enhance our ability to discern normal patterns of PVS beyond those reported at a lower magnetic field (Bouvy et al., 2014, Maclullich et al., 2004, Wuerfel et al., 2008). Bouvy et al. reported normal PVS at 7 T using isotropic spatial resolutions of 0.5 × 0.5 × 0.5 mm3 and 0.7 × 0.7 × 0.7 mm3 with magnetization prepared rapid gradient echo (MPRAGE) and T2-weighted variable flip angle (VFA) turbo spin echo (TSE) sequences, respectively (Bouvy et al., 2014). However, it is unclear if the resolutions employed by Bouvy et al. were sufficient to clearly delineate normal PVS, nor is it clear if the imaging parameters were optimized. Furthermore, quantitative measures of morphological features, such as length or diameter distributions, were not provided. To this end, we carried out both simulation and experimental studies to optimize imaging parameters of the MPRAGE and VFA-TSE sequences (Busse et al., 2006, Mugler and Brookeman, 1990) to maximize contrast-to-noise ratio (CNR) for PVS at 7 T. Specifically, simulations were carried out to study the sequence parameter dependence of PVS to white matter (WM) contrast in T1 and T2-weighted sequences. Subsequently, the sequence with the imaging parameters that provided the highest CNR was employed to image PVS, which in turn enabled segmentation of PVS over the whole imaging volume using a semi-automatic method. Finally, we analyzed the morphological features of extracted PVS, including diameter, length, and volume distributions over the WM and subcortical nuclei including basal ganglia and thalamus.
Section snippets
Methods and materials
A total of 7 healthy volunteers were recruited (ages 21–37). The study was approved by local institutional review board and written consents were obtained from all subjects. All images were acquired on a 7 T Siemens scanner using a 32-channel receive head coil and a single-channel volume transmit coil (Nova Medical, Wilmington, MA). Due to the known B1 inhomogeneity at 7 T (Vaughan et al., 2001), the transmitter voltage was calibrated such that the flip angle (FA) matched with the true FA at the
Simulation
Figs. 2(A) and (B) show examples of the simulated evolution of the echo train signals in PVS and WM for the MPRAGE and TSE sequences, respectively. Here, the MPRAGE was simulated with TI = 1800 ms and FA = 8°. The contrasts calculated as the signal differences between PVS and WM at the echo for ky = 0 are denoted by the vertical lines. It is evident that the maximum contrast in the TSE sequence is ~ 10 times higher than that in the MPRAGE sequence, indicating that TSE sequences are more suitable for
Discussion
In this paper, simulation and experimental evaluations of the CNR of T1-weighted MPRAGE and T2-weighted VFA-TSE sequences for visualization of PVS in the human brain were conducted at 7 T. Both simulation and experimental results show that TSE sequence provides a much higher CNR than MP-RAGE, which can be explained by two factors. First, the ratio of T2 (= 8.0) between PVS and WM is much greater than that of T1 (= 2.6), leading to a greater signal difference between the two tissues with a T2
Conclusions
In conclusion, the T2-weighted VFA-TSE sequence had a much higher CNR for PVS visualization than the T1-weighted MPRAGE sequence. The optimized T2-weighted sequence with TE = 319 ms and 0.4 mm isotropic resolution enabled the visualization of a large number of PVSs in WM and SN in young healthy subjects. A semi-automatic method was developed to segment the PVS and reveal the detailed distributions of the length, volume, and diameter of PVS. Our approach enables accurate characterization of the
References (40)
- et al.
Venous collagenosis and arteriolar tortuosity in leukoaraiosis
J. Neurol. Sci.
(2002) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages
Comput. Biomed. Res.
(1996)- et al.
Immunological and neuropathological significance of the Virchow–Robin space
J. Neurol. Sci.
(1990) Multiecho imaging sequences with low refocusing flip angles
J. Magn. Reson.
(1988)- et al.
Improved optimization for the robust and accurate linear registration and motion correction of brain images
Neuroimage
(2002) - et al.
Myelin water imaging of multiple sclerosis at 7 T: correlations with histopathology
Neuroimage
(2008) - et al.
T2*-based fiber orientation mapping
Neuroimage
(2011) - et al.
MP2RAGE, a self bias-field corrected sequence for improved segmentation and T1-mapping at high field
Neuroimage
(2010) - et al.
Vascular inflammation in cerebral small vessel disease
Neurobiol. Aging
(2012) - et al.
Comparison of physiological noise at 1.5 T, 3 T and 7 T and optimization of fMRI acquisition parameters
Neuroimage
(2005)
Blood–brain barrier and cerebral small vessel disease
J. Neurol. Sci.
User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability
Neuroimage
The in vivo influence of white matter fiber orientation towards B(0) on T2* in the human brain
NMR Biomed.
Arterial pulsation-driven cerebrospinal fluid flow in the perivascular space: a computational model
Comput. Methods Biomech. Biomed. Engin.
Visualization of perivascular spaces and perforating arteries with 7 T magnetic resonance imaging
Investig. Radiol.
Fast spin echo sequences with very long echo trains: design of variable refocusing flip angle schedules and generation of clinical T2 contrast
Magn. Reson. Med.
Enlargement of Virchow–Robin spaces in cranial trauma: literature review
Recenti Prog. Med.
Simplified mathematical description of longitudinal recovery in multiple-echo sequences
Magn. Reson. Med.
Enlarged perivascular spaces on MRI are a feature of cerebral small vessel disease
Stroke
Multiscale vessel enhancement filtering
Cited by (46)
Imaging perivascular space structure and function using brain MRI
2022, NeuroImageCitation Excerpt :Since the PVS volume in healthy individuals is strongly correlated with the brain size (Barisano et al., 2021b), it is recommended to correct for it when implementing statistical models aimed at comparing PVS volumes across different individuals. It is also possible to compute the volume fraction of PVS in each brain region, calculated by dividing the total volume of segmented PVS cluster voxels in the region by the total volume of the region (Barisano et al., 2021b; Zong et al., 2016). The mean cross-sectional diameter is another important PVS feature that can be used to distinguish PVS from other lesions related to cerebral small vessel disease (CSVD) (Wardlaw et al., 2013).
Volumetric distribution of perivascular space in relation to mild cognitive impairment
2021, Neurobiology of AgingCitation Excerpt :To ensure that the results we found in asMTL were not confounded by the presence of white matter hyperintensities, which we excluded from the study, we compared voxels that were segmented as white matter hyperintensities in FLAIR images across CN and MCI and found no significant difference (p = 0.7). It is also fundamental to consider that the MRI-based PVS analysis strongly relies on the image resolution: higher resolution corresponds to increased PVS detection and improved segmentation (Barisano et al., 2018; Madai et al., 2012; Zong et al., 2016). Small PVSs remain still undetected on MRI or are not discernible from noise.
Perivascular Space Imaging at Ultrahigh Field MR Imaging
2021, Magnetic Resonance Imaging Clinics of North AmericaMorphology of perivascular spaces and enclosed blood vessels in young to middle-aged healthy adults at 7T: Dependences on age, brain region, and breathing gas
2020, NeuroImageCitation Excerpt :Therefore, further studies are needed to further clarify the age dependence of the PVSVs. Leveraging the increased signal to noise ratio at 7T, we have previously demonstrated that large numbers (~500) of PVSVs can be visualized in young healthy subjects using ultra-high-resolution transverse relaxation time (T2) weighted (T2w) MRI (Zong et al., 2016). Furthermore, we have developed a deep learning based segmentation method to facilitate automatic delineation of the PVSVs (Lian et al., 2018), allowing more detailed and quantitative characterization of their morphological features.