The effect of noise and depolarization on hyperpolarized tracers perfusion assessment

The effect of imaging pulse sequences on the perfusion model of hyperpolarized tracers and its influence on the quantitative assessment of blood perfusion were simulated and studied. Using three different singular value decomposition (SVD) based regularizations, including truncated SVD and first and second order Tikhonov, perfusion related parameters were calculated and compared. The simulation results revealed that depolarization compensated and uncompensated models underestimate the flow rate by similar amounts. However, the compensated model determines the mean transition time and cerebral blood volume more accurately than the uncompensated model. It was shown that, due to the rapid decay of the effective residue function in hyperpolarized tracers, regularization techniques sensitive to the first derivative of the residue function fail in perfusion parameter estimation