The long-term objective of this work is the development of a general, practical, PET detector that can achieve very good spatial resolution for the imaging situation at hand without a significant compromise in other important properties, specifically, sensitivity, and energy and timing resolutions. The primary goal of this project is to design a thick (25-mm), continuous detector with an independent depth-of-interaction (DOI) measurement. The DOI measurement, besides reducing parallax effect in the scanner, will enable a coarse sampling of the 3D light spread along the depth (z) and a fine sampling in the two transverse directions (x,y). With such a characterization of the 3D light spread, we expect to maintain good spatial resolution in the thick detector. The ease of manufacture and reduced cost, improved packing fraction, and promise of good or better energy and timing resolution compared to pixelated detectors, makes this detector design practical and high performance. The main innovation over previously investigated continuous detectors is the ability to achieve and maintain high spatial resolution in thick detectors over most of the detector FOV, making the detector a practical solution for PET imaging due to its high sensitivity. The high spatial resolution of this detector together with DOI measurement will allow the manufacture of smaller ring diameter scanners which provides increased sensitivity, reduced cost due to less scintillator used, and reduced annihilation photon non-collinearity effect on spatial resolution. Recent work with thin crystals suggests that it is possible to achieve very high spatial resolution in continuous detectors. However, since the light spread in a continuous detector is a function of all three crystal dimensions and leads to degeneracies in the accuracy of the positioning algorithm, it is almost impossible to maintain this spatial resolution over the entire field-of-view (FOV) of detectors that are thick enough to provide the high sensitivity needed for PET. In this work we will establish the use of a lightguide and independent DOI measurement to de-couple the 3D nature of the light spread in a thick detector, thus improving transverse resolution. This work will result in a splitting of the detector into several interaction depth levels, with good transverse resolution within each level. Also, we will investigate through Monte Carlo the ability to match the measured light spread in these detectors. This Monte Carlo will be used to describe the full 3D light spread in the detector that can further improve spatial resolution, and also lead to an easier calibration procedure for the detector. Finally, different crystal materials and surface finishes, as well as varying light sampling techniques, will be investigated through simulations and measurements to achieve good spatial resolution while maintaining the energy and timing resolution in these detectors. Clinically for human imaging, limited spatial resolution of current generation of PET scanners leads to partial volume effects in the measured uptake of small lesions. By improving the spatial resolution to the best possible limit, the overall efficacy of clinical diagnosis will be significantly improved and small lesions that often represent the initial stage of cancer will be characterized early enough to guide the therapy. Hence, thick continuous detectors have the potential to significantly impact the clinical imaging situation and the resultant patient treatment. ? ? ?
| Kaul, M; Surti, S; Karp, J S (2013) Combining Surface Treatments with Shallow Slots to Improve the Spatial Resolution Performance of Continuous, Thick LYSO Detectors for PET. IEEE Trans Nucl Sci 60:44-52 |
| Surti, Suleman; Werner, Matthew E; Karp, Joel S (2010) Imaging studies for evaluating impact of position sampling techniques in PET scanners. IEEE Trans Nucl Sci 57:2497-2503 |
| Surti, S; Karp, J S (2008) Design considerations for a limited angle, dedicated breast, TOF PET scanner. Phys Med Biol 53:2911-21 |
