Optimization of scan parameters in pulmonary partial pressure oxygen measurement by hyperpolarized 3He MRI

The dependence of hyperpolarized (HP) (3)He T(1) on local oxygen concentration provides the basis for measuring the partial pressure of oxygen (pO(2)) and oxygen depletion rate (R) in the lungs. Precise measurements of this type are difficult because the oxygen effect manifests itself through a decay of signal, leading to noisy images at the end of the series. The depolarization caused by RF excitation pulses further complicates the problem. It is therefore important to optimize scan parameters, such as measurement timing and flip angle, to obtain accurate and reproducible measurements. This work presents a new single-acquisition technique in conjunction with the multiple regression fitting method for data evaluation. Analytical expressions for the measurement uncertainties are derived. A total of four types of single-acquisition timing schemes are investigated; simulation shows a large uncertainty variation between these schemes (pO(2): 7.5-30.2%; R: 47.4-173.7%). A basic procedure for optimizing scan parameters is then described. A phantom experiment was conducted to verify the simulation results. Repeated in vivo measurements with the optimal scheme in a rabbit experiment showed that average variation of global mean is 6.2% for pO(2) and 12.0% for R, and that the average variation of percentiles (10th, 25th, 50th, 75th, and 90th) is 8.7% for pO(2) and 19.0% for R.