TOMOGRAPHIC RECONSTRUCTION OF SMALL OBJECTS USING A PRIORI KNOWLEDGE

摘要

Three-dimensional (3D) reconstruction of a cell (1) includes adjusting acurrent set of projection images (8, 9) according to a priori knowledge (24)to produce adjusted projection images (9), for example, based on probabilitymasks and/or Bayesian analysis (16) of multiple similar objects in the samesample (30). A reconstruction algorithm processes the adjusted projectionimages to generate a 3D image (10). The 3D image is further adjusted accordingto the a priori knowledge (24) to generate an adjusted 3D image (11). Criteriafor process completion are applied to determine whether the adjusted 3D imageis adequate (12). Otherwise, a set of pseudo projections are computationallycreated at the same projection angles as the current set of projection images(8, 13) and then compared to the current set of projection images (8) toproduce a set of new projections (14), wherein the new projections are inputagain to the reconstruction algorithm and the steps of the method are repeateduntil the adequacy criteria are met (15).