Abstract #4382

Rapid Hyperpolarized-Gas Lung Imaging Using a Parallel-Spiral Acquisition with BOSCO Reconstruction

Hao Tan¹, Weitian Chen^2,3, Peng Hu^2,4, G. W. Miller⁵, T. A. Altes⁵, J. F. Mata⁵, E. E. de Lange⁵, G. D. Cates^5,6, Ray F. Lee⁷, Craig H. Meyer^2,5, John P. Mugler III^2,5

¹Electrical Engineering, University of Virginia, Charlottesville, VA, USA; ²Biomedical Engineering, University of Virginia, Charlottesville, VA, USA; ³now with GE Healthcare; ⁴now with Beth Israel Deaconess Medical Center; ⁵Department of Radiology, University of Virginia, Charlottesville, VA, USA; ⁶Physics, University of Virginia, Charlottesville, VA, USA; ⁷Princeton Neuroscience Institute, Princeton University, Princeton, NJ, USA

We developed two variable-density parallel-spiral sequences with parallel image reconstruction based on successive convolution operations (BOSCO) reconstruction for rapid lung imaging using hyperpolarized 3He. These sequences yielded images with spatial resolution and image quality comparable to standard GRE images. We also developed the first BOSCO image reconstruction technique for single-shot spiral, using inverse gridding for BOSCO training. Future work will concentrate on optimizing the flip angles to maximize SNR, and on developing a 3D parallel-spiral acquisition.