Introduction to holospectral imaging in nuclear medicine for scatter subtraction

Abstract

An approach to image analysis and processing, called holospectral imaging, is proposed for dealing with Compton scattering contamination in nuclear medicine imaging. The method requires that energy information be available for all detected photons. A set of frames (typically 16) representing the spatial distribution at different energies is then formed. The relationship between these energy frames is analyzed, and the original data is transformed into a series of eigenimages and eigenvalues. In this space it is possible to distinguish the specific contribution to the image of both primary and scattered photons and, in addition, noise. Under the hypothesis that the contribution of the primary photons dominates the image structure, a filtering process can be performed to reduce the scattered contamination. The proportion of scattered information removed by the filtering process is evaluated for all images and depends on the level of residual quantum noise, which is estimated from the size of the smaller eigenvalues. Results indicate a slight increase in the statistical noise but also an increase in contrast and greatly improved ability to quantitate the image.