Scattering Correction of CBCT Projections for the Quality Enhanced CT Reconstruction Using an Efficient Monte Carlo Photon Transport Model

Abstract

In industrial computed tomography, cone beam computed tomography (CBCT) setups are state-of-art because of its capabilities and speed. However, this technique highly suffers from the effect of scatter in comparison to fan beam computed tomography due to the large scanning 2D field of view. CBCT is more preferable to be used since it is considerably faster. In this work a fast Monte Carlo forward projection model based on photon transport has been developed to subtract the scatter from the measured X-ray projections. The ability of this forward model to compute any complex object from CAD models based on meshes polyhedron validate the proposed scatter subtraction algorithm. Depending on the number of photons used, the complexity and the size of the object each X-ray projection is obtained within two minute or less with this model. In addition to the use of the variance reduction techniques for a fast calculation of the scatter, the proposed scatter subtraction algorithm has been further reduced in computation time by the use of smoothing filters, mainly by the use of Savitzky-Golay filter which is a type of a squares polynomial approximation filter, this kind of smoothing plays significant role in the reduction of the computations time.