A review of the image segmentation and registration methods in liver motion correction in C-arm perfusion imaging

The C-arm perfusion imaging is not yet the standard part for liver cancer diagnosis and intraoperative measurements. The number of multiple rotations of C-arm depends on the total angle range resulting in large number of projections. Usually most of the liver disease features can be visible on CT images. However, the breathing and other liver surrounding organs motion during the acquisition can cause artifacts or missing important data on reconstructed CT images. In C-arm systems this is the consequence of the slow rotation. This paper gives an overview of existing motion correction algorithms used in medical imaging. These algorithms are based on image segmentation and registration.