The overall goal of this renewal application is to improve the effectiveness of positron emission tomography (PET) imaging for the diagnosis and treatment of patients with cancer. The proposed research will extend the comparison of the performance of several alternative scanner designs for tumor quantitation and detection tasks commonly performed in our clinical and research PET protocols. The target applications include tumor imaging in the breast and torso using 18F-estradiol, 18FDG, and 11C-thymidine, for tumor detection and staging, therapy planning, and monitoring response to therapy. As proposed in the previous funding period, we have carried out comparisons of 2D and 3D modes of operation of existing tomographs and begun to study the impact of different collimator and scanner geometries on the primary tasks. The initial data on the partial collimation effort (both simulated and measured in our tomographs with modified septa) indicated the potential for significant improvements in overall performance of tomographs and is the major motivation for this renewal application. In this competing application we extend our work on partial collimation (number and size of septa) and different scanner geometries (crystal type, crystal size, ring diameter, and axial field-of-view). We also consider how these changes complement/are complemented by time-of-flight (TOF). The work entails simulation and experimental validation. The proposal includes the development and validation of the appropriate normalization and scatter corrections for the various partial collimation and time-of-flight schemes. Simultaneously, we will simulate a range of partial tomograph geometries and evaluate them using the noise equivalent count (NEC) metric. Those geometries with higher NECs will be selected for further evaluation both with and without TOF, using both quantitative metrics and observer studies. We will include a range of body types and sizes. In all of the proposed effort, the underlying hypothesis is that for large patients the combination of partial collimation and TOF will provide a significant improvement in image quality and quantitation. As part of our previously funded research, a Simulation System for Emission Tomography (SimSET) has been developed. This Monte Carlo-based software package tracks photons through complex heterogeneous objects (patients) and models the effects of collimation and detection. The package has been essential to our efforts and has been adopted by more than 250 other investigators in their own work. Part of our effort will be to continue to improve SimSET with more efficient collimator simulation tools. The SimSET software package is freely available via the Internet.
Showing the most recent 10 out of 32 publications
