Meeting Banner
Abstract #1876

The Mathematics of HYPR

Angel Ramon Pineda1, Atousa Sarcon1, Nasser Abbasi1, Doug Stang1, Siavash Jalal1, Kacie Jacklin1, Reed F. Busse2, Jean H. Brittain2

1Mathematics Department, California State University, Fullerton, CA, USA; 2Applied Science Laboratory, GE Healthcare, Madison, WI, USA


HYPR is a promising new method for time-resolved imaging that uses a time averaged image (composite) to improve images at individual time frames. We show that HYPR and a subsequently described modification by Huang and Wright (HW-HYPR) can be understood mathematically as the first step of iterative methods used in other imaging modalities: HYPR as an approximation to maximum likelihood expectation maximization (MLEM) and HW-HYPR as the multiplicative arithmetic reconstruction technique (MART) used to solve the normal equations, where the composite image serves as an initial estimate prior to iteration.