A uniform and simultaneous algebraic reconstruction technique for X-ray digital tomosynthesis

The conventional visual or optical methods for testing and shape measurement of three dimensional objects have inherent shortcomings such as occlusion and variant surface reflection. In this paper, we propose a three dimensional X-ray imaging method to reconstruct the three dimensional structure of an object. To achieve this, two or more X-ray images projected from different views are required, which are available in the developed digital X-ray imaging system. Once these images are acquired, any arbitrary cross section of an object can be built by computation using the given image set through the digital tomosynthesis method. By integrating a series of cross sections thus obtained, we can build a three dimensional volume of the object. However, the reconstructed three dimensional volume of an object in this way is approximate and not exact to that of the real object due to artifacts, an inherent shortcoming of the digital tomosynthesis method. To overcome the limitation and reconstruct a more accurate volume, we employ an iterative approach known as the uniform and simultaneous algebraic reconstruction technique. In this method, all voxels within the volume are equally weighted to update their internal density values, thereby achieving a uniform convergence property in the entire reconstruction region. The algorithm is simulated on various shapes of objects. Based on these simulation results, the performance of the proposed method is compared with that of the conventional simultaneous algebraic reconstruction technique.