SCATTER COMPENSATION METHODS IN 3D ITERATIVE SPECT RECONSTRUCTION - ASIMULATION STUDY

Effects of different scatter compensation methods incorporated in full y 3D iterative reconstruction are investigated. The methods are: (i) t he inclusion of an 'ideal scatter estimate' (ISE); (ii) like (i) but w ith a noiseless scatter estimate (ISE-NF); (iii) incorporation of scat ter in the point spread function during iterative reconstruction ('ide al scatter model', ISM); (iv) no scatter compensation (NSC); (v) ideal scatter rejection (ISR), as can be approximated by using a camera wit h a perfect energy resolution. The iterative method used was an ordere d subset expectation maximization (OS-EM) algorithm. A cylinder contai ning small cold spheres was used to calculate contrast-to-noise curves . For a brain study, global errors between reconstruction and 'true' d istributions were calculated. Results show that ISR is superior to all other methods. In all cases considered, ISM is superior to ISE and pe rforms approximately as well as (brain study) or better than (cylinder data) ISE-NF. Both ISM and ISE improve contrast-to-noise curves and r educe global errors, compared with NSC. In the case of ISE, blurring o f the scatter estimate with a Gaussian kernel results in slightly redu ced errors in brain studies, especially at low count levels. The optim al Gaussian kernel size is strongly dependent on the noise level.