Three-dimensional whole-brain perfusion quantification using pseudo-continuous arterial spin labeling MRI at multiple post-labeling delays: accounting for both arterial transit time and impulse response function

• Published in: NMR in biomedicine Vol. 27; no. 2; pp. 116 - 128
• Main Authors: Qin, Qin, Huang, Alan J., Hua, Jun, Desmond, John E., Stevens, Robert D., van Zijl, Peter C. M.
• Format: Journal Article
• Language: English
• Published: England Blackwell Publishing Ltd 01.02.2014; Wiley Subscription Services, Inc
• Subjects: Adult; Algorithms; arterial transit time; Blood Flow Velocity - physiology; Brain; Brain - physiology; cerebral blood flow; Cerebrovascular Circulation - physiology; clinical; Computer Simulation; Female; GRASE; human; Humans; Image Interpretation, Computer-Assisted - methods; Imaging, Three-Dimensional - methods; impulse response function; Magnetic Resonance Angiography - methods; Male; Middle Aged; Models, Cardiovascular; PCASL; Pulse Wave Analysis - methods; Reproducibility of Results; Sensitivity and Specificity; Spin Labels; impulse response function; PCASL; clinical; GRASE; cerebral blood flow; brain; human; arterial transit time
• ISSN: 0952-3480; 1099-1492; 1099-1492
• DOI: 10.1002/nbm.3040

Measurement of the cerebral blood flow (CBF) with whole‐brain coverage is challenging in terms of both acquisition and quantitative analysis. In order to fit arterial spin labeling‐based perfusion kinetic curves, an empirical three‐parameter model which characterizes the effective impulse response function (IRF) is introduced, which allows the determination of CBF, the arterial transit time (ATT) and T1,eff. The accuracy and precision of the proposed model were compared with those of more complicated models with four or five parameters through Monte Carlo simulations. Pseudo‐continuous arterial spin labeling images were acquired on a clinical 3‐T scanner in 10 normal volunteers using a three‐dimensional multi‐shot gradient and spin echo scheme at multiple post‐labeling delays to sample the kinetic curves. Voxel‐wise fitting was performed using the three‐parameter model and other models that contain two, four or five unknown parameters. For the two‐parameter model, T1,eff values close to tissue and blood were assumed separately. Standard statistical analysis was conducted to compare these fitting models in various brain regions. The fitted results indicated that: (i) the estimated CBF values using the two‐parameter model show appreciable dependence on the assumed T1,eff values; (ii) the proposed three‐parameter model achieves the optimal balance between the goodness of fit and model complexity when compared among the models with explicit IRF fitting; (iii) both the two‐parameter model using fixed blood T1 values for T1,eff and the three‐parameter model provide reasonable fitting results. Using the proposed three‐parameter model, the estimated CBF (46 ± 14 mL/100 g/min) and ATT (1.4 ± 0.3 s) values averaged from different brain regions are close to the literature reports; the estimated T1,eff values (1.9 ± 0.4 s) are higher than the tissue T1 values, possibly reflecting a contribution from the microvascular arterial blood compartment. Arterial spin labeling (ASL) signals are not only proportional to the cerebral blood flow (CBF), but also depend on the arterial input function characterized by the arterial transit time (ATT) and the impulse response function characterized by T1,eff. Both simulation and experimental data using our multi‐delay ASL protocol demonstrated a feasible three‐parameter model that renders the simultaneous extraction of CBF, ATT and T1,eff. The data analysis on healthy subjects indicates that both the two‐parameter model using fixed blood T1 for T1,eff and the proposed three‐parameter model provide reasonable fitting results.