MONTE-CARLO ANALYSIS OF CAMERA-INDUCED SPECTRAL CONTAMINATION FOR DIFFERENT PRIMARY ENERGIES

The problem of scatter contamination has received considerable attention from many research groups during the past decades. In nuclear medicine, scatter can dramatically affect image contrast and activity quantification. Several scatter analysis techniques have been based on theoretical models of the contaminated and/or the uncontaminated energy spectra. This paper shows that the imaging instrumentation itself may limit the effectiveness of strategies for compensating for object scatter because of the overlapping of 'good' and 'bad' events throughout the energy spectrum. We have found that the following factors are mostly responsible for the spectral contamination: (i) scatter in the object, (ii) fractional energy deposition or multiple interactions in the detector and (iii) scatter in the collimator. For instance, in a typical gamma camera configuration, close to 10% of the peak energy events at 150 keV are mispositioned while more than 6% of the non-peak events are placed in the correct 2 mm pixel. The same ratios rapidly increase to 27% and 23%, respectively, for 250 keV photons. We also show that a collimator that alters only minimally the spectrum at 150 keV could become one of the most important contamination factors at 250 keV. Possible effects on several scatter compensation techniques are considered.