Cone-beam computed tomography (CT) has some advantages that it can make a three-dimensional (3D) image in shorter time and with higher resolution in longitudinal direction of body than the helical CT scanner. However, because large amount of scatter radiation is included in 2D projection data, the quality of reconstructed image is worse and cupping artifacts appear in CT images. In this study, we have tried to improve the quality of reconstructed images and obtained the following results :
1.Using Monte Carlo simulation, we analyzed quantitatively the intensity distribution of scatter radiation on 2D projection data. From these simulation data, we have evaluated the dependence of scatter radiation on 3D reconstructed images.
2.Simulation results showed that intensity distribution of incident x-ray were controlled primary radiation on detector to be uniform, the scattered radiation included in projection data were decrease. Then, we reconstructed 3D CT images using a bow-tie flitter, which controlled the incident x-ray distribution and confirmed that the image quality of CT images improved.
3.We have improved the quality of CT images using image processing techniques, such as multiresolution analysis of wavelet transform, antidiffusion equation model and software correction with scattered distribution. These techniques were very useful to improve the quality of reconstructed image degraded by scatter radiation. Especially, usage of the weighting factors which were obtained from multiresolution analysis, we could reconstruct the CT images without any extra computing time and improve the image quality.
We Confirm our results are effective to make the cone-beam CT scanner in practical use.