Data weighted vs. non-data weighted dual energy reconstructions for X-ray tomography

X-ray transmission tomography is useful for estimation of attenuation correction for simultaneously obtained emission tomography images. Moreover, X-ray attenuation coefficients can be combined with emission measurements in order to create combined anatomical-functional images. It is imperative, then, that as low X-ray dosage as possible be used. At those fluxes implementation of iterative algorithms becomes desirable. However, that was not the case in clinical practice where X-ray fluxes used are high enough and noniterative algorithms give good results. In order to use the acquired data in a statistically efficient manner an appropriate model has to be employed. Also an efficient algorithm has to be used to solve the objective function. In this paper it is shown that at low fluxes an appropriate weighting becomes a crucial part of the model. Results are presented first for the single energy case and then for the dual energy case.