A Four-Dimensional Image Reconstruction Framework for PET under Arbitrary Geometries.

摘要

Positron Emission Tomography (PET) is a functional imaging modality with applications ranging from the treatment of cancer, studying neurological diseases and disease models. Virtual-Pinhole PET technology improves the image quality in terms of resolution and contrast recovery. The technology calls for having a detector with smaller crystals placed near a region of interest in a conventional whole-body PET scanner. The improvement is from the higher spatial sampling of the imaging area near the detector. A prototype half-ring PET insert built to study head-and-neck cancer imaging was extended to breast cancer imaging. We have built a prototype half-ring PET insert for head-and-neck cancer imaging applications. In the first half of this work, we extend the use of the insert to breast imaging and show that such a system provides high resolution images of breast and axillary lymph nodes while maintaining the full imaging field of view capability of a clinical PET scanner.;We are focused on designing unconventional PET geometries for specific applications. A general purpose 4D PET reconstruction framework was created to estimate the radionuclide uptake in the subject. Quantitative estimation in PET requires precise modeling of PET physics. Data acquired in a PET scanner is well modeled as a Poisson counting process. Reconstruction given the forward model is implemented using MAP-OSEM. The framework is capable of reconstructing PET data under arbitrary position of the detector elements and different crystal sizes. A novel symmetry finding algorithm is created to reduce the system matrix size, without loss of resolution. The framework motivates investigation into different PET system geometries for different applications, as well as optimizing the design of PET systems. A generalized normalization procedure was developed to model unknown components. The programs are parallelized using OpenMP and MPI to run on small workstations as well as super-computing clusters. The performance of our reconstruction framework is presented through four novel and unconventional PET systems, each designed specifically for a different geometry. The Virtual-Pinhole half-ring system is a half-ring insert integrated into a Siemens Biograph-40, for head and neck imaging. The Flat-panel system is a modular insert system integrated into the Biograph-40, designed for breast cancer imaging. The MicroInsert II is the second generation full ring insert device, integrated into the MicroPET scanner to improve the resolution and contrast recovery of the MicroPET scanner. The Plant PET system is a PET system designed to image plants vertically, and integrated into a plant growth chamber. The improvement in speed/memory from symmetry finding is as high as a factor of 50 in some cases. Further improvements to the framework and state of the field are also discussed.