Method for reconstructing three-dimensional image data set from two-dimensional projection images of target area for patient neck bone, involves adjusting model parameter correspondence with object that is visible to reference image

摘要

The method involves utilizing a start model (8) that describes a part of a target area for initialization of reconstruction process and/or for correction of a raw image data. A model parameter of a statistical object model (5) is determined by a main component analysis from basic image data sets (2) of an anatomical object (3) i.e. neck bone, for determination of the start model for the object of the target area. The model parameter is adjusted correspondence with the object that is visible to a reference image i.e. projection image (6a), of the currently captured target area. An independent claim is also included for a computing device.