Cerebral Blood Flow Measurements by Magnetic Resonance Imaging Bolus Tracking: Comparison with [¹⁵O]H₂O Positron Emission Tomography in Humans

Abstract

Summary: In six young, healthy volunteers, a novel method to determine cerebral blood flow (CBF) using magnetic reso nance (MR) bolus tracking was compared with [,sO]H20 pos itron emission tomography (PET). The method yielded para metric CBF images with tissue contrast in good agreement withRecent results indicate that it may be possible to mea sure CBF by dynamic magnetic resonance imaging (MRI) of paramagnetic contrast agent bolus passage (0stergaard et aI., 1996a). Because of the complexity of susceptibility contrast, this technique initially only al lowed determination of relative flow rates. In a prelimi nary study in six normal volunteers, the mean gray to white flow ratio was found to be in good agreement with PET literature values for age-matched subjects (0ster gaard et aI., 1996b). In a recent animal hypercapnia study (0stergaard et aI., 1998), an approach was introduced to

Partial volume effect (PVE) on the arterial input function (AIF) in T1‐weighted perfusion imaging and limitations of the multiplicative rescaling approach

“…20. The procedure F s3c is generally nonlinear, and hence rescaling by multiplication of concentration curves as employed previously (6,(13)(14)(15)(16)(17) is not equivalent to multiplication of signal curves and may distort the AIF (4). We instead apply the multiplication to the MR signal (Eq.…”

“…Furthermore, such an approach increases quantification errors due to delay and dispersion of the bolus (11), which on the contrary motivates the use of a local AIF (12). Thus several studies, either based on dynamic susceptibility contrast (DSC) perfusion imaging (6,13,(15)(16)(17)(18)(19) or dynamic contrast-enhanced (DCE) perfusion imaging (14), have compensated the PVE by a multiplicative rescaling. The rescaling is either applied to the AIF (6,(13)(14)(15)(16)(17) or the CBF estimate (18,19), and is often based on normalization to a venous outflow function (VOF) (14,16 -18).…”

Principles of cerebral perfusion imaging by bolus tracking

“…In an attempt to obtain absolute flow values from EPI experiments, Østergaard et al assumed proportionality between the area of the AIF and the injected contrast dose, using water clearance PET as a calibration method. This approach provided reproducible absolute CBF measurements in animal hypercapnia studies (6) and in humans (43). However, this approach may be too crude to allow general use in patients with severe cardiac or cerebrovascular disease.…”

Arterial input functions for dynamic susceptibility contrast MRI: Requirements and signal options

“…The AIF signal should be on the same scale as the tissue curve, or it should be possible to place the AIF signal on the same scale as the tissue curve by applying a relative scale factor (similar to the approach used in Ref. 55). While the use of a ⌬R2* AIF with a ⌬R2* tissue curve may seem appealing because the same signal type is used, both the ⌬R2* and the ⌬ AIF curves require scaling to place them on the same concentration scale as the tissue ⌬R2* data.…”