System optics in at least in one of backprojection and forward projection for model-based iterative reconstruction

摘要

Iterative reconstruction (IR) algorithms are advantageous over standard filtered backprojection (FBP) algorithms by improving resolution and noise performance. In this regard, model-based IR algorithms (MBIR) have been developed to incorporate accurate system models into IR and result in a better image quality than IR algorithms without a system model. System optics are included in both forward and backprojection (IRSOM-FPBP).

主权项

1. A method of reconstructing an image using a scanner, comprising:
acquiring data for an object; estimating an image based upon the data; iteratively performing steps of:
forward projecting to generate projection data using a first predetermined one of system optics models and non-system optics models, wherein the system optics models include at least a voxel model, a focal spot model, a detector element model and a rotation blur model; andback projecting to update an image using a second predetermined one of the system optics models.