Parameter Estimation Approach to T1 and T2 Mapping in Phase-Cycled bSSFP Imaging

Abstract

Relaxation times T1 and T2 can be estimated simultaneously from a phase cycled balanced steady state free precession (bSSFP) sequence, when the bSSFP signal is considered to have an elliptical signal model. Parameters of the model can be found by solving an ellipse specific linear least squares. Hence, estimation of the model parameters directly relies on the resulting ellipse fits. Therefore in the presence of highly deviated noise, estimations diverge from the true parameter values. In order to tackle this problem, a patch-based approach is presented in this study. With this approach, data points of the neighbor voxels are used to decrease the effect of noise on estimations. This method is compared with original ellipse fitting approach in terms of mean absolute percentage error and structural similarity.