SCATTER CORRECTION IN THE TRANSAXIAL SLICES OF A WHOLE-BODY POSITRON EMISSION TOMOGRAPH

A procedure for scatter correction in a whole-body positron emission tomograph (PET) is presented. The method is based on measured scatter distributions of line sources. It extends the method originally developed by Bergstrom et al for brain scanners to large objects. Scatter distributions were measured with line sources at different positions in cylindrical water containers with diameters ranging from 20 to 35 cm. The projection data were approximated by a monoexponential function, the scatter distribution function f. The parameters of this function depended on the projected distance of the source from the centre and on the diameter of the phantom. For scatter correction, the scatter distributions in emission scans were calculated by integral transformation of the measured projection data with the spatially variant scatter distribution function, and then subtracted from the measured projection data. The method was tested using cylindrical phantoms of different diameters and the EEC body phantom with arms, both with uniform activity distributions. In the cylindrical phantoms scatter fractions between 9.8% (10 cm diameter) and 14.1% (35 cm diameter) were found. The reduction of scatter contributions in the reconstructed images of uniform sources was assessed using inactive water cylinders with 5 cm and 8 cm diameters inserted in the uniform source phantoms. In the centre of the phantom with 35 cm diameter, for example, the scatter contamination was reduced from 25% to 4.1% by the scatter correction. With scatter correction the calibration of the scanner was independent of the object size, whereas without correction the measured activity concentrations varied by 14% when going from 10 cm to 35 cm diameter. The results of the study demonstrate that the application of this scatter correction markedly improves image uniformity and activity determination.