Timing and Energy Response of Six Prototype Scintillators

Transcription

1 Timing and Energy Response of Six Prototype Scintillators CCM Kyba 1, J Glodo 2, EVD van Loef 2, JS Karp 1, KS Shah 2 1 University of Pennsylvania 2 Radiation Monitoring Devices SCINT 2007 June 7, 2007

2 Motivation (1) t 1 t 2 Standard PET uses lineof-response (LOR) to determine activity distribution The use of time of flight (TOF) information reduces correlations TOF reconstruction algorithms take advantage of local positioning to improve sensitivity

3 Clinical Advantage of TOF-PET Example: Colon cancer images for a heavy weight patient CT non-tof TOF Improvement in lesion detectability with TOF Improving timing capabilities will extend TOF benefits to all patients

4 New scintillators are interesting for future TOF- PET scanners Understanding differences between benchtop measurements and scanner performance Energy and timing measurements are necessary for the planning of future scanners Motivation (2)

5 Scintillator Samples from RMD YI 3 (2% Ce) LuGdI 3 (2% Ce) LuI 3 (2, 5, 10% Ce) CeBrCl 3 Samples are hygroscopic, and packaged inside glass window Variations in packing affect light output LaBr 3 (5% Ce) crystal from Saint-Gobain used as reference

6 Timing Measurement Setup ADCs LaBr 3 H4998 Gate TDCs CAMAC 22 Na sample Coincidence H4998 PC CFDs Delay

7 Timing and Energy Comparison Sample Resp (%) E FWHM (%) T FWHM (ps) LaBr 3 (5%Ce) LuI 3 (2% Ce) CeBrCl LuGdI 3 (2% Ce) YI 3 (2% Ce) LuI 3 (5% Ce) LuI 3 (10% Ce)

8 Energy Linearity Setup ADC Also test 133 Ba, 137 Cs, and 57 Co Gate CAMAC 22 Na sample H4998 CFD PC

9 Fitting Multiple 133 Ba Peaks Counts 276keV 303keV 356keV 384keV Channel Requires good energy resolution Easy with LaBr 3 (4.4% FWHM), very difficult beyond 10% E res If peaks are not distinguishable used ~294 and 356keV

10 LaBr 3 Energy Proportionality ADC Channel 133 Ba 137 Cs 22 Na Eight energy points used in fit Data agrees with a linear fit within uncertainties, but the intercept is not at zero 57 Co E (kev)

11 Energy Proportionality (ADC Chan)/Fit LaBr 3 Uncertainty due to changes in pedestal, larger fraction at lower ADC values LaBr 3 is proportional in the range studied E (kev)

12 Energy Proportionality (ADC Chan)/Fit LaBr 3 LYSO Uncertainty due to changes in pedestal, larger fraction at lower ADC values LaBr 3 is proportional in the range studied LYSO noticeably non-proportional E (kev)

13 Energy Proportionality (ADC Chan)/Fit LaBr 3 LuI 3 E (kev) Uncertainty due to changes in pedestal, larger fraction at lower ADC values LaBr 3 is proportional in the range studied LYSO noticeably non-proportional LuI 3 in between

14 Energy Proportionality (ADC Chan)/Fit LaBr 3 CeBrCl 3 E (kev) Uncertainty due to changes in pedestal, larger fraction at lower ADC values LaBr 3 is proportional in the range studied LYSO noticeably non-proportional LuI 3 in between CeBrCl 3 in between

15 Energy Proportionality (ADC Chan)/Fit LaBr 3 LYSO LuI 3 CeBrCl 3 E (kev) Uncertainty due to changes in pedestal, larger fraction at lower ADC values LaBr 3 is proportional in the range studied LYSO noticeably non-proportional LuI 3 in between CeBrCl 3 in between

16 Further Considerations for TOF-PET Intrinsic properties Density Z Light output Fast decay time Physical effects Stopping power Photo-fraction Energy resolution Crystal determination Lower dead time Timing resolution Scanner performance Sensitivity Spatial resolution Scatter rejection Spatial resolution Higher count rates TOF positioning The image quality, lesion detectability, and quantification possible on a clinical scanner are dependent upon each of these properties

17 Summary LuI 3 and CeBrCl 3 are already interesting possibilities for future TOF-PET scanners LuI 3 has a higher stopping power than LaBr 3 The PMT used in this study was not well matched to the frequency spectrum of LuI 3. A better match would lead to increased light detection/energy resolution Crystal development and testing continues Multiple samples and long crystals Development of detector arrays