New reconstruction and analysis for coded-aperture imaging

In this paper, a new reconstruction algorithm and an analytic approach of collimator optimization for coded-aperture imaging were demonstrated. Different from those mentioned in present literatures, the calculations in the above methods are in frequency domain. We proved that decoding method is a special case of our reconstruction algorithm. In the analysis of code-aperture imaging properties, spectrum of Poisson noise on reconstructed image were related directly with transfer function of the collimator in frequency domain. A principle of optimizing the design of coded-aperture collimator was given based on maximizing SNR in both projection data and reconstructed images. Though the above methods were generally concerning 2D imaging, they can also be expanded to the 3D reconstruction and analysis.