Many motion correction methods have been proposed in the past, for either respiratory motion or body motion in PET. However, there has been no unified approach that can simultaneously correct for both motions. In this study, Dr. Lu, Prinicipal Investigator, proposes to develop algorithms to correct both body and respiratory motions simultaneously for PET/CT. Specifically, a data-driven method (i.e., based on PET raw data itself) will be developed to detect both body and respiratory motions for both single-bed and whole-body PET, followed by event-by-event non-rigid motion corrected reconstruction. Dr. John Onofrey will be working with Dr. Lu on motion estimation optimization using non-rigid image registration algorithms.
Submitted by Angel Machon on August 16, 2019