Filtered backprojection using algebraic filters; application to biomedical micro-CT data

For computerized tomography (CT) imaging in (bio)medical applications, radiation dose reduction is extremely important. This can be achieved simply by reducing the number of projection images taken. In order to obtain accurate reconstructions from few projections, however, common reconstruction techniques are not sufficient. Algebraic reconstruction methods (ARMs) are often more suited, but inflict a much higher computational burden. In this work, a recently proposed method is applied to biomedical μCT, in which the benefits of ARMs are combined with the computational efficiency of the common Filtered Backprojection (FBP) algorithm. Our experimental results demonstrate that this approach yields reconstructed images highly similar to those obtained by an ARM, while maintaining the favorable computational efficiency of FBP.