Iterative Energy Spectrum Deconvolution (IESD): a new technique for significantly improving the effective energy resolution of gamma camera

Energy windows used clinically in nuclear medicine (20-30%) accept a significant proportion of Compton scattered photons to highly degrade image contrast. We developed an Iterative Energy Spectrum Deconvolution (IESD) algorithm for improving the energy resolution of gamma camera to reduce the impact of photon scatter on radionuclide images. Using only those deconvoluted energy components near the photopeak, we exclude scattered photons to form scatter corrected projections. All phantom studies reconstructed with attenuation correction and IESD correction were shown significant improved contrast as cardiac and hepatic activity separation. Contrast of presenting defect region was also significantly enhanced in the phantom and patient studies. Computational times for energy deconvolution times were within clinically acceptable norms (∼14 minutes/64 projections, 2 GHz, 100 iters). The computational requirements of this technique make it feasible for routine clinical use.