基于Bayesian非局部先验的低剂量X-射线CT成像

摘要

Objective To reduce the radiation dose delivered to the patients during the application of computed tomography (CT). Methods CT images can be easily degraded by the quantum noise under low X-ray dose scan protocols. Statistical reconstructions outperform the traditional filtered back-projection (KBP) reconstructions by accurately modeling the scan system and the measurement statistics. This paper aims to improve the CT reconstruction using a new nonlocal prior statistical reconstruction approach. Results Compared to traditional reconstruction approaches, the proposed nonlocal prior can impose an effective regularization for reconstructions by exploiting the image global information adaptively. Conclusion Kxperimentalion validates that the proposed CT reconstructions have excellent performance with low-dose scan protocols.%目的 降低病人在CT扫描中所受的辐射剂量.方法 在低剂量的扫描条件下,由于有效投影信息的缺少,CT图像易受到量子噪声的影响.本文提出了一种新的非局部先验Bayesian重建算法来提高CT图像重建质最.结果 与传统的图像重建方法比较,本文提出的重建方法,能够利用目标图像中更多形态结构的全局信息来构建先验项,从而克服了传统图像重建信息局限性的缺点.结论 试验证明本文所提出的CT图像重建方法在低剂量扫描条件下具有很好的表现.