Purpose: In pinhole SPECT, attenuation of the photon flux on trajectories between source and pinholes affects quantitative accuracy of reconstructed images. Previously we introduced iterative methods that compensate for image degrading effects of detector and pinhole blurring, pinhole sensitivity and scatter for multi-pinhole SPECT. The aim of this paper is (1) to investigate the accuracy of the Chang algorithm in rodents and (2) to present a practical Chang-based method using body outline contours obtained with optical cameras.
Methods: Here we develop and experimentally validate a practical method for attenuation correction based on a Chang first-order method. This approach has the advantage that it is employed after, and therefore independently from, iterative reconstruction. Therefore, no new system matrix has to be calculated for each specific animal. Experiments with phantoms and animals were performed with a high-resolution focusing multi-pinhole SPECT system (U-SPECT-II, MILabs, The Netherlands). This SPECT system provides three additional optical camera images of the animal for each SPECT scan from which the animal contour can be estimated.
Results: Phantom experiments demonstrated that an average quantification error of -18.7% was reduced to -1.7% when both window-based scatter correction and Chang correction based on the body outline from optical images were applied. Without scatter and attenuation correction, quantification errors in a sacrificed rat containing sources with known activity ranged from -23.6 to -9.3%. These errors were reduced to values between -6.3 and +4.3% (with an average magnitude of 2.1%) after applying scatter and Chang attenuation correction.
Conclusion: We conclude that the modified Chang correction based on body contour combined with window-based scatter correction is a practical method for obtaining small-animal SPECT images with high quantitative accuracy.