a new approach for scatter removal and attenuation compensation from spect/ct images

objective(s): in spect, the sinogram contains scatter and lack of attenuated counts that degrade the reconstructed image quality and quantity. many techniques for attenuation and scatter correction have been proposed. an acceptable method of correction is to incorporate effects into an iterative statistical reconstruction. here, we propose new maximum likelihood expectation maximization (mlem) formula to correct scattering and attenuating photons during reconstruction.   materials and methods: in this work, scatters are estimated through klein-nishina formula in all iterations and ct images are used for accurate attenuation correction. reconstructed images resulted from different mlem reconstruction formula have been compared considering profile agreement, contrast, mean square error, signal-to-noise ratio, contrast-to-noise ratio and computational time. results: the proposed formula has a good profile agreement, increased contrast, signal-to-noise (snr) & contrast-to-noise ratio (cnr), computational time and decreased mean square error (mse) compared with uncorrected images and/or images from conventional formula. conclusion: in conclusion, by applying the proposed formula we were able to correct attenuation and scatter via mlem and improve the image quality, which is a necessary step for both qualitative and quantitative spect images.