We present a novel method for joint reconstruction of both image and motion in positron-emission-tomography (PET). Most of nowadays motion compensation methods consist of two completely separated steps: (i) motion estimation and (ii) image estimation. A major drawback of these methods lies in the motion estimation step, since it is completely based on the usually noisy individually reconstructed gates. As we show in a simulation study, a joint reconstruction approach alleviates this drawback and results in both visually and quantitatively better image quality. We attribute these results to the fact that for motion estimation always the currently best available image estimate is used and vice versa. Additionally, results for dual respiratory and cardiac gated patient data are presented.
展开▼
