Investigation of image noise in cone-beam CT imaging due to photon counting statistics with the Feldkamp algorithm by computer simulations

Abstract

Cone beam CT (CBCT) imaging technique has found many applications in medical imaging in recent years since it provides three-dimensional image reconstruction with low contrast detectability and high resolution. In CBCT systems, image noise caused by photon counting statistics plays an important role in image quality by affecting low contrast detection and patient dose. In this paper, we studied image noise caused by photon counting statistics with the Feldkamp's cone beam reconstruction algorithm. A noise model based on photon counting detector was developed to provide the noise distribution in 3D image space and the quantitative relationships between image noise level and various CT parameters, including the exposure level, the number of projections, the reconstruction filter and the detector pixel size. Computer simulations of a spherical water phantom were conducted to test the theoretical model. This work provides the noise spatial distribution in cone-beam CT and its relationship with CBCT parameters. It can also provide local noise information for adaptive de-noising in the future.