Practical method of quantifying cerebral blood flow using arterial spin labeling. Numerical simulation and correlation of pulsed ASL and cold xenon-CT CBF

Currently arterial spin labeling (ASL) is not used in clinical practice even though various studies have been carried out on quantifying cerebral blood flow (CBF) by magnetic resonance (MR) perfusion imaging using ASL. In the following study, we evaluated the relationship between ASL signals and cerebral blood flow both numerically and experimentally. When evaluating the relationship numerically we used simulation based on the two-compartment model. We evaluated experimentally by using the correlation of pulsed ASL (PASL) signals with the CBF measured by cold xenon-CT (Xe-CT) in the human brain. Based on the results, we propose a practical linear scaling method for quantifying CBF values using a single TI pulsed ASL image. Numerical simulation using the two-compartment model, which was considered to model reality more closely than the conventional single-compartment model, showed that the ASL signal was correlated almost linearly in the CBF range (<100[ml/100 cc/min]). The ASL signal was also stronger in variations of model parameters, such as the relaxation time in extravascular space (T1e), intravascular volume (CBV), and the brain-blood partition coefficient, than when the conventional single-compartment model was used. A linear correlation between the ASL signal and the absolute CBF value was also found in the PASL and Xe-CT experiments. This indicates that it may be possible to quantify CBF values using only a single TI pulsed ASL image. The proposed method requires reference CBF data (Xe-CT, SPECT, PET data or flow phantom data) or both a control image and conversion table obtained by simulation calculations. Although the ASL signals depend on the MRI hardware and imaging conditions, once the scaling coefficient K is calculated, the ASL signals can be converted to corresponding CBF values. Further studies are required in patients with ischemic conditions because the arterial transit delay time cannot be ignored in these patients. The proposed method can become clinically useful if its applicability is extended to such patients. (author)