Reconstruction of dynamic objects in computerized tomography

Reconstruction of dynamic objects in computerized tomography

One basic assumption in classical computerized tomography is that the object does not change during the data acquisition process. In many applications, this assumption does not hold, for example due to respiratory and cardiac motion in medical imaging. The application of standard reconstruction methods leads to serious motion artifacts in the images. Hence, the algorithms have to take the dynamics of the investigated object into account. To obtain an adequate reconstruction, some a priori information of the motion is required. Therefore, we consider a certain type of movements and develop respective algorithms for the dynamic problem, using the method of the approximate inverse. An algorithm to determine the parameters of the motion from the measured data is proposed as well. Numerical examples illustrate the advantages of the new method.