| 摘要 |
<p>A GPU in this PET device performs processing in four steps (S1 through S4). Specifically, said GPU performs a conversion step (step S2) in which list-mode data is converted to projection data and an estimation step (step S3) in which missing-region data is estimated on the basis of said projection data, and said missing-region data is used as projection data used in an acquisition step (step S4). Converting the list-mode data to projection data in the conversion step (step S2) reduces the spatial resolution (image quality) of the reconstructed images (tomographic images), but since only part of the missing-region data is in the form of projection data, there is no reduction in the actual quality of the data. The projection data can thus be used to compensate for missing data, making it possible to reduce artifacts and improve image quality relative to traditional list-mode reconstruction.</p> |
