4D PET with the quad-HIDAC: development of dynamic list-mode EM image reconstruction

The sub-millimeter imaging capability of the quad-HIDAC , combined with its good count rate performance, offers great potential for biomedical research. This paper presents a methodology for spatio-temporal (4D) list-mode image reconstruction with this scanner. The technique is developed from the 3D list-mode based expectation-maximization (EM) algorithm. Using measured data from both a cylindrical phantom and a mouse study, approaches to reconstructing dynamic image sequences were developed, evaluated and compared in terms of resolution and noise, and different types of time-activity curve were produced and evaluated. The effect of altering the various reconstruction parameters is considered, including the effect of post-reconstruction smoothing, and the image initialization for each time frame reconstruction. Based on these results, it is possible to achieve high image quality with good temporal resolution, arising from the excellent spatial resolution and noise properties characteristic of static quad-HIDAC imaging.