Hemodynamic Parameter Estimation From Magnetic Resonance Perfusion Imaging With the Tikhonov Regularization Method

In the brain, T2-weighted dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) enables the measurements of hemodynamic parameters, such as cerebral blood flow (CBF) and cerebral blood volume. Accurately characterizing the tissue residue function in DSC-MRI is of crucial importance to the quantification of cerebral hemodynamics. The estimation of the tissue residue function is an inverse problem, and one of the approaches is through the deconvolution. In this paper, Tikhonov regularization is used to reconstruct the residue function with smooth constraints. The influences of pertinent factors, such as signal-to-noise ratio (SNR) and tracer delay on the reconstruction, are analyzed in detail. The simulation results show that the SNR and the tracer delay have little influence on the estimation of CBF. Therefore, the Tikhonov regularization method can accurately estimate the CBF with confidence.