Research on the Mixed Noise Model of Low-Photon-Count XLCT

Résumé fourni par la source

X-ray luminescence computed tomography (XLCT) is a novel optical molecular imaging method, which has broad application prospects in aspects such as disease diagnosis, new drug research and development, and curative effect monitoring. However, limited by the probe concentration, luminescence performance, and X-ray dose, the number of X-ray excited photons is currently relatively low, and it is vulnerable to projection noise, which has become a bottleneck restricting the in-vivo application of XLCT. In the XLCT projection images, the XLCT projection noise includes dark current noise, readout noise, and statistical noise caused by weak photon signals. Among them, both the readout noise and the dark current noise are independent of the intensity of the input optical signal and are the noises that naturally exist under dark conditions. The statistical noise caused by weak photon signals is an inherent noise generated due to the input of the optical signal. In this paper, theoretical modeling and experimental verification studies are carried out on the distribution of projection noise in low-photon-count XLCT. The influence of the exposure time on the background noise under dark conditions was evaluated, and the distribution of the detected photons in CB-XLCT was statistically analyzed according to the experimental conditions. Through the statistical analysis of the CB-XLCT projection images, it provides certain guidance for image denoising and improving the imaging quality, and lays a theoretical foundation for subsequent CBXLCT imaging based on the statistical noise model.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.