TH-A-18C-05: Scatter Reduction and Correction for Dual-Source CBCT Using the Synchronized Moving Grid (SMOG) System




to implement the newly developed synchronized moving grid (SMOG) system on a prototype dual-source CBCT system to evaluate its efficacy for scatter reduction and correction under various settings, including different phantom size, grid ratio and image acquisition modes.


A 1D grid was designed for each tube in our dual-source CBCT system. Projections were acquired with the grid at different complementary positions from each scan angle and merged for reconstruction. The scatter signal was measured from the grid blocked area and the scatter distribution was estimated by using a cubic spline interpolated/extrapolated in each projection for scatter correction. The projections without grid were also acquired as the comparing group. Three sets of images were reconstructed from projections: a) without grid, b) with grid but without scatter correction and c) with grid and with scatter correction to evaluate CNR (contrast-to-noise ratio) and CT number linearity enhancement. The efficacy of the SMOG system was evaluated using CAT phantoms with different diameters (15cm, 20cm and 30cm), grids (grid ratios of 1:1 and 2:1) and acquisition modes (simultaneous: two tubes firing at the same time and sequential: only one tube firing in one rotation).


That scatter artifacts were substantially reduced with our scatter correction algorithm. CNR, degraded by the cross scatter in dual-source system, were increased by 74%, 32% and 28% for phantom size of 15cm, 20cm and 30cm respectively after using scatter correction with 1:1 grid. And in 20 cm phantom case with simultaneous acquisition, the CT number linearity improved from 0.9969 to 0.9997. Higher grid ratio (more blocked area) resulted in better scatter artifact removal and CNR improvement at the cost of complexity and increased exposure.


The SMOG system can effectively reduce the scatter artifacts and enhance CNR and CT number linearity for the dual-source CBCT system.

This work was supported by NIH grant R01CA166948.