2386

Characterization of Cerebrospinal Fluid Using ultra-high field MRI

Tiago Martins¹, Tales Santini¹, Minjie Wu¹, Kristine Wilckens¹, Davneet Minhas¹, James W. Ibinson¹, Howard J. Aizenstein¹, and Tamer S. Ibrahim¹
¹University of Pittsburgh, Pittsburgh, PA, United States

Synopsis

In this work, we investigated the oscillations on the flow cerebrospinal fluid in the human brain using ultra-high field magnetic resonance imaging with a homogeneous head coil. Acquisition and analysis of images from 5 human volunteers results in identification of different frequency bands around 0.3, 0.8, 1.2, 2.3 and 3.4Hz. The analysis of the frequency spectrum and spatial localization of the signal yield results that could be correlated with physiological processes and CSF clearances.

Introduction

The high signal-to-noise ratio (SNR) of 7 Tesla magnetic resonance imaging (MRI) and a coil with good field homogeneity allow studies to perform analysis of blood and cerebrospinal fluid (CSF) flow within the brain¹. Using fast echo-planar imaging (EPI) and physiological acquisitions, it is possible to view the CSF MR signal in real-time. Analysis of the CSF flow can improve the study of brain diseases and sleep^2,3. This work presents the findings regarding the CSF oscillation patters obtained from 5 human subjects.

Methods

To study the CSF flow, we acquired fast EPI sequences of 5 healthy volunteers were scanned using a 7T Siemens MAGNETON and the current 16Tx/32Rx Tic Tac Toe head coil design⁴. A fast EPI sequence (Figure 1) with resolution of 2x2x2mm was acquired using TE of 20ms and TR of 152ms for subject 1 and 155ms for the others. Due to constraints of the system, the acquisition was broken into separate slabs of 3 axial slices each. A total of 15 slabs were acquired for subject 1 and 19 slabs for the other subjects, therefore the total number of slices are 45 for subject 1 and 57 for the other subjects. This way it was possible to obtain a field of view large enough to cover the whole brain. The positioning of the acquisition was done to obtain data from below the cerebellum from CSF coming from the spine.
The EPI acquisition yields a real-time visualization of CSF flow (Figure 1). Electrocardiogram measurements were collected for correlation between the MRI signals and the physiological activities (Figure 2). The processing and analysis of the signal intensity was performed using MATLAB, and ANTs for a processing pipeline was develop for distortion correction, denoising and bias correction of the EPI datasets. For each subject, the frequency spectrum was calculated for 5 arbitrary points to validate findings across different regions of the brain (Figure 3).
From the frequency spectrum, 4 bands of approximately 0.24Hz each were chosen for a spatial analysis. A mask as created for spatial localization of each frequency band. The amplitudes of the points of the 3D space were processed for each frequency within the desired band to produce a mask of the approximate region that was later overlayed on top of the original EPI acquisition for anatomical reference (Figure 4).
For subject 1, a faster EPI sequence with same resolution (2x2x2mm) was acquired using TE of 20ms and TR of 51ms for a single slice. With this acquisition, it was possible to obtain a much larger frequency spectrum up to 9.8Hz (Figure 5) when compared to the 3.23Hz of the other acquisitions.

Results

The frequency spectrum as calculated for 5 arbitrary and different points for each subject. By, analyzing the spectrum for all 5 subjects, 4 bands can be visually identified (Figure 3) for having higher signal amplitude. The center frequencies for these bands are approximately 0.3Hz, 0.8Hz, 1.2Hz and 2.3Hz.
The spatial localization of these frequencies can be seen in Figure 4. Lower frequencies can be seen where a larger volume of CSF flows (the main ventricles and the outside of the brain). Higher frequencies can be seen close to the cortex and at the ventricles.
The larger frequency spectrum (Figure 5) shows all 4 previously identified frequency bands plus another frequency band around 3.4Hz.

Discussion and Conclusion

The raw visualization of the real-time signal (Figure 1) can be helpful for diagnosis and analysis of CSF motion. The flow of CSF within the ventricles and in the outside of the brain changes the signal intensity and can be clearly seen (Figure 1c). The frequency analysis for multiple volunteers shows a consistent result and enables the speculation of the origin of each frequency band.
These finding could be correlated with physiology of the human body as well as CSF clearance. For example, for the first band could be result of a breathing process propagating through the CSF (one breath every 4 seconds or 0.25Hz)^5,6 and the third band could be correlated with cardiac motion (approximately 72 bpm)^6,7. The other bands could be harmonics or other physiological processes, including the higher frequency band of 3.4Hz.
Future work includes acquisition of faster EPI sequences to increase the span of the frequency spectrum, measurement of physiological signals in scanner, correlation between the physiological signals and the MRI signals and study of possible correlation with amyloid clearance in the brain, brain waves and sleep⁸. These correlations will give us better understanding of brain diseases and aging. A more advanced processing pipeline will also be used to result in clearer images and less noisy data.

Acknowledgements

This work was supported by NIH R01MH111265 and R01AG063525. The author Tiago Martins was partially supported by the CAPES Foundation, Ministry of Education of Brazil, 88881.128222/2016-01. This research was also supported in part by the University of Pittsburgh Center for Research Computing (CRC) through the resources provided.

References

[1]A. Scouten and R. T. Constable, “VASO-based calculations of CBV change: accounting for the dynamic CSF volume,” Magn Reson Med, vol. 59, no. 2, pp. 308–315, Feb. 2008, doi: 10.1002/mrm.21427.

[2] N. E. Fultz et al., “Coupled electrophysiological, hemodynamic, and cerebrospinal fluid oscillations in human sleep,” Science, vol. 366, no. 6465, pp. 628–631, Nov. 2019, doi: 10.1126/science.aax5440.

[3]L. Xie et al., “Sleep Drives Metabolite Clearance from the Adult Brain,” Science, vol. 342, no. 6156, pp. 373–377, Oct. 2013, doi: 10.1126/science.1241224.

[4] N. Krishnamurthy et al., “Computational and experimental evaluation of the Tic-Tac-Toe RF coil for 7 Tesla MRI,” PLOS ONE, vol. 14, no. 1, p. e0209663, Jan. 2019, doi: 10.1371/journal.pone.0209663.

[5]V. Kiviniemi et al., “Ultra-fast magnetic resonance encephalography of physiological brain activity – Glymphatic pulsation mechanisms?,” J Cereb Blood Flow Metab, vol. 36, no. 6, pp. 1033–1045, Jun. 2016, doi: 10.1177/0271678X15622047.

[6]B. A. Telfer et al., “Estimating Sedentary Breathing Rate from Chest-Worn Accelerometry From Free-Living Data,” in 2020 42nd Annual International Conference of the IEEE Engineering in Medicine Biology Society (EMBC), Jul. 2020, pp. 4636–4639, doi: 10.1109/EMBC44109.2020.9175669.

[7]C. Schiweck, D. Piette, D. Berckmans, S. Claes, and E. Vrieze, “Heart rate and high frequency heart rate variability during stress as biomarker for clinical depression. A systematic review,” Psychological Medicine, vol. 49, no. 2, pp. 200–211, Jan. 2019, doi: 10.1017/S0033291718001988.

[8]M. Massimini, R. Huber, F. Ferrarelli, S. Hill, and G. Tononi, “The Sleep Slow Oscillation as a Traveling Wave,” J. Neurosci., vol. 24, no. 31, pp. 6862–6870, Aug. 2004, doi: 10.1523/JNEUROSCI.1318-04.2004.

Figures

Figure 1: Fast EPI acquisition (TR=100ms) showing signal changes due to CSF flow; axial slices with resolution of 1.53 x 1.53 x 3mm and sagittal slice with resolution of 1.5 x 1.5 x 4.36mm; a) bottom axial slice; b) top axial slice; c) sagittal slice.

Figure 2: Data acquisition along the cardiac cycle. Echo-planar imaging acquisition performed with concurrent physiological measurement of electrocardiogram in the 7T scanner. Signals are temporally aligned using external trigger signal.

Figure 3: Frequency spectrum of the signal intensity in 5 arbitrary points within the brain for 5 different subjects (a-e).

Figure 4: Frequency spectrum with spatial localization of the signal from 4 different frequency bands. For each band, the example show a spatial coverage of the signal ranging from the bottom of the brain/lower cerebellum on the left to the center of the brain on the right. The bandwidth for each band is 0.24Hz. The center frequency is approximately a) 0.3Hz, b) 0.8Hz, c) 1.2Hz, d) 2.3Hz. The acquisition was done using an EPI sequence with TR=152ms with 15 slabs of 3 slices each for a total of 45 slices.

Figure 5: Frequency spectrum with 5 highlighted band. The center frequency for each band is approximately 0.3, 0.8, 1.2, 2.3 and 3.4Hz respectively. The acquisition was done using an EPI sequence single slice with TR=51ms.

Proc. Intl. Soc. Mag. Reson. Med. 29 (2021)

2386