The goal of this proposal is to develop the full potential of PET for whole body imaging. Whole body imaging with PET is in the early stages of development, and a number of areas requiring improvement are apparent. (1) In a whole body scan, the time available for collecting data for any portion of the body is limited by considerations of patient comfort and/or motion and the biological and physical life time of the radiopharmaceutical. Therefore, the events collected per image slice are very limited. (2) Because of the practical difficulties of obtaining attenuation correction measurements for the whole body, and since calculated attenuation correction techniques have not been developed for all parts of the body, the quantitative aspects that are possible in PET measurements have largely been ignored in the initial development of whole body PET imaging. (3) Because of the large data set, the data processing burden for the PET system is excessive. These problems are addressed in this proposal with a combination of new hardware and new software techniques for collection and processing of whole body PET data. (1) Image data will be collected in a true 3-D format by removing interplane septa and collecting all coincidence combinations. This can provide on the order of a 7 fold increase in gross sensitivity. Images will be reconstructed with a true 3-D algorithm that correctly uses essentially all the collected data. (2) Since the short time available for individual scan segments will generally preclude measured attenuation correction for whole body imaging, a variety of approximate attenuation correction techniques using only emission data or short transmission scans will be developed and evaluated. (3) Scatter will be significantly higher without interplane septa and a much larger correction will be required. Various analytic scatter correction techniques will be implemented and evaluated in this environment, and in addition, a new scatter correction technique which makes use of a second energy window on the PET system will also be implemented and evaluated. (4) The problem of excessive computation times for reconstruction and other aspects of data processing will be addressed by implementing a cost effective parallel processing system that can be added to most PET computer system with little CPU burden except for I/O. The culmination of the work described in this proposal will be in its application to the new PET systems which will have axial fields of view of 15 to 20 cm. These configurations will provide an additional factor of 2 to 4 in sensitivity for 3-D data collection, giving improvements of factors of 10 to 20 in sensitivity over the 2-D configurations currently found in PET systems.
Showing the most recent 10 out of 17 publications
