Image noise properties in circular sinusoid cone-beam CT

Recently, a reduced circular sinusoid scan was proposed in which data are collected only over an open portion of the full circular sinusoid trajectory. In general, for reconstructing the same ROI image, the reduced-scan approach uses less cone-beam data than what required for a full circular sinusoid scan. It is of practical significance to investigate and evaluate the numerical and noise properties of the image reconstruction for reduced circular sinusoid trajectories. In this work, we carry out studies on the noise properties in images for full and reduced circular sinusoid trajectories reconstructed by use of the recently developed full- scan and reduced-scan FBP algorithms. It can be observed that the spatially dependent weighting factors in the FBP algorithms can lead to non-uniform image variances.