Exploring the need for a preload on the estimation of permeability, vessel radius, and relative cerebral blood volume in MR vascular fingerprinting-based dynamic susceptibility contrast perfusion imaging.

PURPOSE

Dynamic susceptibility contrast (DSC) MRI is commonly part of the clinical brain tumor imaging protocol. Usually, a preload of contrast agent is administered to minimize contrast-leakage T effects. However, recent studies have indicated that with adaptation of scan parameters (in particular, low flip angle), a preload is not required. For advanced approaches, like MR vascular fingerprinting (MRVF), which is based on combining gradient-echo and spin-echo readouts, excitation flip angles cannot easily be reduced, but postprocessing might alleviate leakage effects. In this study, the aim was to verify whether DSC-MRVF can be performed without preload.

METHODS

A previously established simulation model was used to create a dictionary of DSC-MRVF signals resulting from a double-bolus injection with varying input values for permeability, vessel radius, and relative cerebral blood volume (rCBV). A sensitivity analysis based on a Cramer Rao Lower Bound approach was performed, from which the logarithmic ratio of standard deviations for parameter estimation was derived. A ratio above zero corresponded with a better determination without preload. Two brain tumor patients were included for validation.

RESULTS

The results show that the standard deviation for determining permeability was lower without preload, whereas for determining rCBV, the opposite result was found. For vessel radius, the standard deviation was similar with and without preload. The in vivo data reasonably agreed with the simulation results.

CONCLUSION

This work has shown the potential of DSC-MRVF with a single bolus injection for determining vessel radius and permeability, although for rCBV, a preload is still advisable.