INTRODUCTION

Cerebrovascular reactivity (CVR) mapping with BOLD fMRI is an important indicator of hemodynamic function (1) , particularly in small cortical blood vessels (2), and is increasingly utilized in research and clinical applications. There is evidence that the baseline vascular state is coupled with CVR. A positive correlation between baseline CBF and CVR, both between- and within-subjects has been reported (3,4). However, there are no studies showing the correlation between CVR with other physiological parameters such as oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen (CMRO₂). In this study, these baseline parameters were measured with MRI and the correlations between CVR and baseline CBF, OEF and CMRO₂ were investigated.

METHODS

Six healthy volunteers (2 female, 4 male, aged 33 ± 6 years) were examined prospectively using a 20-channel head coil on a 3T scanner (Magnetom Prisma, Siemens Healthineers, Erlangen, Germany). All participants provided written informed consent, and the study was approved by the institutional ethics committee. To measure CVR, breath-hold (BH) respiratory challenges were integrated into the brain imaging protocol: for block-designed BH tasks, 110 measurements were obtained, which include five and a half BH/FB (free breathing) cycles with 20 measurements (32s) per cycle and 10 measurements (17s) per half cycle. To measure CBF, a pseudo-continuous arterial spin labeling (pCASL) sequence with 3D gradients and spin-echo imaging (GRASE) readout was applied. Presaturation before labeling and background suppression during postlabeling delay (PLD) were added. Oxygen extraction fraction (OEF) was based on a gradient-echo sampling of spin-echo (GESSE) pulse sequence. Cerebral metabolic rate of oxygen (CMRO₂) was calculated from CBF and OEF. Specific sequence parameters and data analysis are as follows.
CVR: FOV=220×220 mm², matrix size=64×64×28, resolution=3.4×3.4×3.5 mm³, slice gap=0.7 mm, in-plane iPAT factor=2, multiband factor=2, bandwidth = 1776 Hz/px, TE= 27.08 ms, TR=1700 ms. In conventional CVR analysis, measurements of end-tidal CO₂ (Et-CO₂) are required and used in a linear regression equation (5). Without the measurement of Et-CO₂, the CVR was estimated by replacing Et-CO₂ with the mean signal of gray matter in our case.
CBF: FOV=220×220 mm², matrix size=64×64×24, resolution=3.4×3.4×5 mm³, slice and plane PF=6/8, slice oversampling=16.7%, FA=120°, segments=2, Bandwidth = 2298 Hz/px, labeling duration = 1.8s, PLD=1.8 ms, TE=17.18 ms, TR=4500 ms. This sequence had 20 tag and 20 control volumes and one M₀ volume, for a total scan time of approximately 7 min. CBF was calculated using ASLtbx (6). A labeling efficiency of 0.86 was assumed in the calculation.
OEF: FOV=256×192 mm², partial Fourier=6/8, matrix size=128×96×30, resolution=2×2×3 mm3, slice gap = 0.9 mm, TE=51ms; TR=105ms, number of total echoes=64, number of echoes before echo center=20, averages=3, acquisition time was about 10 min. For OEF analysis a feedforward ANN (artificial neural network) was used because of its high capability for nonlinear regression problems (7). The ANN consisted of a 64-dimensional input layer, two hidden layer (32 and 10-dimensional), and a 4-dimensional output layer and was implemented using the Neural Network Toolbox provided by Matlab. The ANN was trained based on the full quantitative BOLD (qBOLD) model. Thereby, artificial BOLD signals with a known ground truth were simulated for plausible ranges of the input variables S_SE, R₂, λ, and OEF. To reduce overfitting artifacts, noise was added to the simulated BOLD signals before the actual ANN training was initialized.
CMRO₂: After smoothing and normalization of CBF and OEF into MNI space, CMRO₂ was calculated as (8):
$$CMRO_{2}=CBF\cdot OEF \cdot [H]_{a}$$
The oxygenated heme molar concentration in arterioles [H]_a was assumed to be 7.377 μmol/mL.

RESULTS

The individual CVR, CBF and OEF could be calculated and respective maps are shown in Fig. 1. Normalized and averaged maps are presented in Fig. 2. A positive correlation between CBF in gray matter and white matter and CVR across subjects was found. Conversely, negative correlations between OEF in gray matter and white matter and CVR was found. Positive correlations between CMRO₂ in gray matter and white matter and CVR across subjects were also found. Specific results are presented in Fig. 3.

DISCUSSION

Gradient-echo (GE) BOLD signal originates from the venous concentration of deoxyhemoglobin (9). Accordingly, we measured CVR based on GE BOLD. Comparing to previous studies, a positive correlation between CBF and CVR was confirmed in our case (3,4). The observed negative correlation between OEF and CVR may be due to venous oxygenation: Y_v is the venous oxygenation, defined as the fraction of oxyhemoglobin in the venous blood, and OEF=1-Y_v: This means that if OEF increases, Y_v decreases and CVR decreases.

CONCLUSION

This study presents the correlation of CVR in healthy brain with baseline CBF, OEF, and CMRO₂.

Acknowledgements

No acknowledgement found.