Progressive updating approach for volumetric CT image reconstruction

Abstract

In overlapping reconstruction for multi-slide spiral CT, adjacent slices of an object typically only have small differences. Based on this property, a progressive updating approach for volumetric ct image reconstruction includes a method and apparatus to use the well-known ordered subset expectation maximization (OSEM) formula or other iterative algorithms for spiral CT overlapping reconstruction. The imaging geometry is assumed to be single-slice helical/spiral scanning. To start with, a complete set of projections is synthesized via linear interpolation of data associated with opposite rays for reconstruction of the first slice. Then, new data are incrementally utilized for reconstruction of subsequent slices by updating the previous slice using the OSEM approach. To overcome accumulative errors, traditional reconstruction from a complete set of projections is performed when it is necessary as is determined by dynamically monitoring the relevant errors.

Dependencies are inferred from the claim text.