Seminar

Medical imaging has been an effective tool for diagnosing diseases including cancer and coronary artery disease, but there are still challenges such as reducing motion artifacts, minimizing toxic radiation dose, and reducing scanning time. These challenges reduce the SNR of the measurements. Model-based image reconstruction and motion estimation can efficiently improve image quality from low SNR measurements. In this model-based approach, constructing regularizers is one of the most critical factors impacting image quality. This talk presents two different regularization designs for motion-compensated image reconstruction (MCIR) methods, which incorporate motion information in image reconstruction to reduce motion artifacts.

 First of all, we investigated methods for motion regularization. Recently, there has been much research on regularizing nonrigid deformations with diffeomorphic motion priors. Conventional methods that enforce deformations to be locally invertible require high computational complexity and large memory. We developed a sufficient condition that guarantees the local invertibility of B-spline deformation and proposed a simple regularizer based on that sufficient condition. This motion regularizer was applied to the motion correction using prototype simultaneous PET-MR. We estimated deformable motion from simultaneously acquired tagged MR volumes and incorporated it into the system matrix of unregularized OSEM algorithm for list-mode PET. We demonstrated the improvement of image quality and detection task performance with deformable phantom, rabbit, and small non-human primate studies.

 Secondly, we studied the spatial resolution properties of MCIR methods and shows that nonrigid local motion can lead to non-uniform and anisotropic spatial resolution for conventional quadratic regularizers. We proposed novel spatial regularization design methods using an “analytical approach” for three different MCIR methods that account for known nonrigid motion. We develop MCIR regularizers that provide approximately uniform and isotropic spatial resolution and that match a user-specified target spatial resolution. 2D PET simulations demonstrate the performance and benefits of the proposed spatial regularization methods.