Medical software control quality using the 3D Mojette projector

The goal of this paper is to provide a tool that allows to assess a set of 2D projection data from a 3D object which can be either real or synthetic data. However, the generated 2D projection set is not sampled onto a classic orthogonal grid but uses a regular grid depending on the discrete angle of projection and the 3D orthogonal grid. This allows a representation of the set of projections that can easily be described in spline spaces. The subsequent projections set is used after an interpolation scheme to compare (in the projection space) the adequation between the original dataset and the obtained reconstruction. These measures are performed from a 3D multiresolution stack in the case of 3D PET projector. Its direct use for the digital radiography reprojection control quality assessment is finally exposed.