Towards Very High Net Acceleration Factors in Hyperpolarized 3He Human Lung Parallel Imaging using SPIRiT

Martin H. Deppe¹, Salma Ajraoui¹, Jim M. Wild¹

Auto-calibration is essential in Parallel Imaging of the human lung using hyperpolarized noble gases due to limitations of breath-hold reproducibility and gas consumption, resulting in a scan time overhead due to the acquisition of auto-calibration lines. This work presents results on Parallel Imaging of hyperpolarized 3He in human lungs using the SPIRiT reconstruction method, showing that this approach maintains image quality better for high net acceleration factors than the commonly used GRAPPA method. As SPIRiT is an iterative optimization method, the inclusion of random sampling and L1-penalization (Compressed Sensing) is straightforward.