1 Nuclear Medicine SPECT and PET 2 a good book! SR Cherry, JA Sorenson, ME Phelps Physics in Nuclear Medicine Saunders, 2012 3 projection tracer 4 PET SPECT CT y E (x)dx L e ( )d L y T I I 0 e ( )d L y E (x) L e ( )d x d dx 5 reconstruction 6 Metabolic images, characteristics depend on tracer specificity tracer sensitivity tracer sensitivity detection system detection system 7 Medical physicist diagnosticsdiagnostics: QC camera hardware & software Image formation, reconstruction Image analysis (dosimetry) therapytherapy dosimetry researchresearch 8 Radionuclides 9 - emission 99 Mo 99m Tc 99 Tc - emission (66 hours) isomeric transition (6 hours) keV 10 Electron capture 11 Electron capture (2.83 days) 0.0 MeV EC MeV MeV 0.0 MeV (stable) 11 22 RadiationfreqMeV 1 ce-K, ce-L 1, 2 ce-K, K 1 X-ray Auger-KLL 12 Positron or + emission 511 keV 11 C 180 o + - positronium 13 Radioactivity 1 mCi = 37 MBq = 37 x 10 6 events per s 14 Poisson noise Chance of measuring n photons when are expected : Poisson distribution resembles Gaussian 15 Poisson noise 16 Poisson noise SNR = n 1 Poisson( 1 ) n 2 Poisson( 2 ) n 1 + n 2 Poisson( ) 17 photon-electron interactions 18 Photon-electron interactionsTissue photo-electric Compton pair production Detectors 19 attenuation scatter photoelectric effect 20 water 21 Energy loss due to Compton scatter -10% E E 22 attenuation 23 attenuation Single photon Positron N(a) N(b)a b ab c 24 attenuation 25 Data acquisition 26 Scintillation time 27 Scintillation crystals NaI(Tl)BGOLSOGSOLaBr:Ce Photons/keV decay time [ns] keV [/cm] wave length melting point transparency, ease of use... 28 Photomultiplier tube 29 Detector design Single crystalmulti- crystal 30 Position and energy measurement output current Electronics X Y x,y,z 31 Multiple events output current Electronics x,y,z All wrong 32 Intrinsic resolution Electronics x,y,z FWHM Collimator Source X Y NaI(Tl): 4 mm 33 Multidetector crystal 34 Expensive alternatives APD: avalanche photo diode diode in reverse mode replaces PMT, much smaller, low voltage works in high magnetic field Cd Zn Te detectors direct detection of high energy photons excellent energy resolution high stopping power (similar to NaI(Tl)) 35 Partial volume effect 36 Partial volume constant activity big pixels 37 Partial volume constant concentration finite resolution perfect resolution finite resolution Recovery Spill-over 38 Collimation 39 PET SPECT CT y E (x)dx L e ( )d L y T I I 0 e ( )d L y E (x) L e ( )d x d dx 40 Collimator LensCollimator 41 Collimators ParallelFanbeam Cone beamPinhole 42 Collimator PSF FWHM position counts 43 Collimator sensitivity a T R r H S r PSF(r) Sens pMol! 44 Collimator sensitivity a T H FWHM 45 d Electronic collimation r x r psf x x d 46 PET lines of response 47 PET sensitivity R d sensitivity in center: 48 PET resolution 511 KeV 0.3 o 2.5 mm (for 1 m FOV) maxmaxmean Mevmmmm 11 C N O F Ga Rb C N O F Ga Rb 49 Coincidence detection TrueScatter SingleRandom 50 PET septa T d D r trues: scatters: singles: randoms: efficiency: time window: 51 2D and 3D PET trues: scatters: singles: randoms: N rings 52 Compton scatter and energy windowing 53 Collimator, scatter, attenuation 54 Energy of scattered photons Measured Simulated primary scattered Counts 55 Scatter window subtraction Simulated primary scattered Energy KeV Counts Scatter PSF is energy dependent 56 Triple Energy Window Simulated primary scattered Energy KeV Counts C2 C1 C3 Corrected counts = C1 - filter(C2 + C3) 57 TEW, 201 Tl cardiac phantom PeakCorrected Lower window Higher window global scale individual scale joint scale 58 Energy resolution Energy KeV Counts SPECT, SPECT, fwhm 10 % PET BGO, PET BGO, fwhm 20 % PET LSO, GSO, LYSO PET LSO, GSO, LYSO... between 10 and 20% Bentourkia, IEEE TMI 1999 59 Model based scatter correction Ollinger 1996, 3D PET S A B 60 PET scatter PSF Measured Monte Carlo Ollinger high sampling Ollinger low sampling Ollinger, Phys Med Biol 1996 61 Activity outside FOVS A B 62 CorrectionsCrystal front end electronics Collimator Computer 63 Linearity correction X Y X + X Y + Y Triad XLT 24 Detector 64 Energy correction 65 Uniformity correction E + L + Flood correctie Energy correction No correction E + Linearity divide by flood source image 66 Uniformity correction non-uniformity due to non-linearity dead PMT 67 head1head2 h1 h2 h1 h2h1 h2 h1 h2h1 h2 Tc-99m MDP bone spect on dual head camera broken PMT 68 e.cam, detector 2 Broken PMT affecting high voltage 69 detector 1detector 2 70 Dead time True count rate Measured count rate 500,000 cps = 700 ns Front end: Data processing: or 71 72 Randoms correction in PET delayed window time Det 1 Det 2 time prompt: true or random random!