Examples of GATE simulation Yun Dong Different systems

Different systems available in GATE • • • Scanner: most generic system; Cylindrical PET:

Example One • System geometry: 3 Rings, 820 mm; each ring has 48 blocks;

Source: 1, 000 Becquerel; back-to-back 511 ke. V gamma ray; uniform distribution in phantom;

Single events event. ID | rs. ID | cry. ID | T(s) | E(Ke.

Coincidence Events event. ID | rs. ID | cry. ID | T(s) | E(Ke.

Example Two Our question: Good energy resolution and bad timing resolution vs. Bad energy

No blurring single events event. ID | rs. ID | cry. ID | T(s)

To evaluate the trade-off between ER and TR, analysis of NECR curve is our

Conclusion Estimation of scanner performance GATE Simulation of PET Imaging Reconstruction New system design

Slides: 11

Download presentation

Examples of GATE simulation Yun Dong

Different systems available in GATE • • • Scanner: most generic system; Cylindrical PET: cylindrical geometry; CPET: simulate C-PET Plus, Philips Medical Systems; SPECT Head: model SPECT dedicated scanners; ECAT: simulate ECAT of CPS Innovations;

Example One • System geometry: 3 Rings, 820 mm; each ring has 48 blocks; each block has 8 8 crystals with 0. 3 mm gap; each crystal is 6. 45 25 mm LSO; Thus, there are totally 24 slices. • Phantom: cylinder, 200 mm, 190 mm length; water

Source: 1, 000 Becquerel; back-to-back 511 ke. V gamma ray; uniform distribution in phantom; Digitizer: Energy Resolution: 15%@511 ke. V Energy Window: 350~650 ke. V Timing Resolution: 2 ns Coincidence Window: 12 ns

Coincidence Events event. ID | rs. ID | cry. ID | T(s) | E(Ke. V) | src. X(mm) | src. Y | src. Z | glb. X 115029 7 43 2. 708248320306 e-03 489. 7 -30. 0 188. 6 0. 5 257. 8 115029 24 36 2. 708249588752 e-03 519. 5 -30. 0 188. 6 0. 5 -420. 2. . . 250179 19 34 5. 858310339553 e-03 446. 0 53. 5 126. 2 22. 1 -324. 1 250179 2 30 5. 858311948285 e-03 423. 5 53. 5 126. 2 22. 1 404. 5. . . 217636 22 19 5. 102768816098 e-03 506. 0 -163. 6 98. 5 -4. 3 -398. 5 217637 39 35 5. 102780608077 e-03 566. 4 45. 2 -153. 6 20. 6 158. 5. . . What we can do with coincidence events? Estimation of performance of system: sensitivity; scatter fraction (SF), S/(T+S); NECR curve: T*T/(T+S+R) Reconstruction: conventional reconstruction via position; reconstruction w/ TOF; reconstruction w/ depth of interaction; Lots of other things | glb. Y 330. 5 -0. 4 | glb. Z | TSR 8. 2 1. 8 T 258. 9 2. 7 126. 3 -0. 9 S 108. 5 -12. 8 -385. 2 3. 5 R

Example Two Our question: Good energy resolution and bad timing resolution vs. Bad energy resolution and good timing resolution Because of many kinds of simulation need to run, we choose in GATE simulation: Turn off the energy resolution Turn off the energy window Turn off the timing resolution Turn off the coincidences window

No blurring single events event. ID | rs. ID | cry. ID | T(s) | E(Ke. V) | src. X(mm) | src. Y 25 31 43 4. 015558794387 e-06 511. 0 -107. 5 -149. 5 640 12 45 1. 255510308222 e-04 511. 0 22. 6 152. 2 883 13 33 1. 636130650978 e-04 511. 0 31. 4 193. 5 1754 30 36 3. 378387639478 e-04 469. 0 -143. 0 -112. 4 …… | src. Z | glb. X 3. 3 -255. 0 2. 4 -11. 6 2. 8 -39. 6 3. 2 -297. 5 | glb. Y | -328. 7 428. 9 415. 2 -305. 5 glb. Z | Scatter 12. 6 0 7. 7 0 2. 5 0 6. 1 1 After we get enough no blurring single events, we use C code to add energy resolution, energy window, timing resolution to the each no blurring single event. Then by the choice of coincidences window, we acquire the coincidence events by ourselves. It is similar with GATE but we can adjust easily the ER, EW, TR and CW followed what we need. That saves time.

To evaluate the trade-off between ER and TR, analysis of NECR curve is our first step doing with TOF PET GATE simulation. The next step we are doing now is to reconstruct images and evaluate the image quality. Until that, we could answer our question.

Conclusion Estimation of scanner performance GATE Simulation of PET Imaging Reconstruction New system design

Thanks