There are many obstacles facing the development of quantitative perfusion techniques in MR. Not the least of which is the inability to dynamically measure the absolute contrast agent concentration ([CA]) quickly and reliably throughout the organ of interest during a bolus injection protocol. We have developed a technique which addresses this need by adapting an older R1 mapping method for dynamic 3D R1 measurement. Since R1 is linearly related to the [CA], a dynamic data set of R1volumes may be substituted for [CA] and absolute perfusion parameters calculated from it.

A pre-contrast scan is first used to determine M0 on a voxel-by-voxel basis throughout the volume of interest using a variable flip angle method. We then inject contrast and scan dynamically as it passes through the vasculature using a Driven Equilibrium Single Pulse Observation of T1 method. By using the ratio of the dynamic signal values and M0 we can solve for R1. From this we obtain an arterial input and tissue residue functions. Using these in a deconvolution algorithm we may extract an impulse residue function, the height of which is the blood flow and the area being the blood volume. Calculating the ratio of these gives us the mean transit time.

In figures 2 and 3 below we have a cerebral blood flow and blood volume map calculated from R1 data. The side scales range from 0-100 ml/min/100g and 0-5 ml/100g respectively.

The flow in the gray matter was measured at 74 ml/min/100g. We also have a clearly defined basal ganglia in the center of the brain with a blood flow of around 42 ml/min/100g These values correlate reasonably well with those quoted in the literature. One of the primary objectives at this point is to optimize the M0 acquisition. This may involve an improved flip angle selection along with attempting to optimize the signal to noise ratio.