Abstract
It is a considerable issue that how to use the functional images for early detecting breast cancer. In this thesis, the feasibility of positron emission mammography (PEM) in detecting breast tumors were investigated with a series of phantom experiments using rotational PET imager developed by Institute of Nuclear Energy Research. The PEM system understudy only uses two opposite BGO detectors, which separated by 220mm. And we employed MLEM algorithm to reconstruct the three-dimensional images. The locations and sizes of the hot spots in breast phantom can be determined from the reconstructed images. The minimum detectable sphere diameter is 2mm. At the target-to-background concentration ratio 5:1, the 10mm size sphere can be visible. And the source out of FOV dose not influence the capability of determining the locations and sizes. In our research, calculation shows that it is possible detecting 2mm size sphere with 22.2kBq in 600 seconds and spending 12 seconds to reconstruction (10 iterations).