Efficiency of antiscatter grids for flat-detector CT

Flat-panel detector CT (FD-CT) scanners offer large volume coverage, but as a consequence are more susceptible to scatter artifacts than standard clinical CT scanners with smaller cone angles. FD-CT scanners can employ antiscatter grids as a scatter rejection technique. We evaluated three standard fluoroscopic antiscatter grids for two different field sizes with respect to scatter suppression efficiency and image quality improvement. The evaluations included simulations and measurements. Regarding the simulation a hybrid model combining deterministic and Monte Carlo (MC) calculations was used combined with an analytical calculation of grid transmission. The scatter-to-primary ratio (SPR) was measured using an adapted collimator technique in order to validate our simulations. The SPR obtained by simulations and measurements with and without antiscatter grids were in agreement typically within 10%. The employment of a grid does not generally provide a significant improvement of the signal-to-noise ratio (SNR). Antiscatter grids led to a significant reduction of cupping artifacts in all cases. There is a trade-off between the SNR and the reduction of the scatter intensity described by the signal-to-noise improvement factor (SNRif). For low- or medium-scatter conditions the increase in noise caused by the reduced primary transmission through the grid has to be compensated by a higher exposure. For high scatter conditions SNRif is significantly greater than 1; i.e. a decrease of dose of up to 50% can be reached.