Synthesis of angiographic images using iterative approximation

摘要: In coronary x-ray angiographies, the vessels supplying heart are imaged in a number of states uniquely determined by combination respiratory intake and contraction patient. The angiographic frames one sequence represent not all possible combinations respiration contraction. A couple applications need a continuous dense sampling state-space given by two axes 'respiration' 'contraction', e.g. background removal or motion-compensated catheter navigation. We present novel method interpolating above twodimensional phase-space based on pairs with similar contraction, but different status. First hypothetical model motion is formulated, rigid transformation translation. Then the parameters that transform single frame into another one status calculated for of frames. An iterative approach used to reconstruct generalized function from transformation parameters pairs. Using this function, arbitrary can be generated. It is shown synthesized angiographies closely match real acquired at same of contraction status.