动态双能区能谱CT成像理论和方法研究

Compared with traditional X-ray CT, spectral CT has the obvious advantage in in the material identification which has been considered as one of the developing trends of CT techniques. However, as the key device the current photon counting detector still has some problems. The complexity of ASIC limits the number of energy bins, and multi-energy-bin design causes the reduction of the data signal-to-noise ratio in each energy bin. In order to overcoming these disadvantages, we proposed a new design of dynamic dual-energy photon counting detector. It has only two dynamic high-/low- energy bins to collect all of the photon in the spectrum. It may obtain as many energy bins as we want by changing the energy threshold of each detector and/or at different acquisition times during a spectral CT scan. This new dynamic dual-energy spectral CT may have higher both energy and spatial resolution, and even lower cost. Aiming to the imaging problems of this dynamic dual-energy spectral CT, this project will study its physical mechanism and mathematical model, spectral CT image reconstruction method and materials decomposition method. Then, the above theories and methods will be validated by both Monte Carlo simulation and real data experiments on our X-ray imaging platform. As a totally new spectral CT design, once we verify the feasibility of its imaging theory and method, it will have important impact on the developments of X-ray detector and spectral CT.

相比传统X射线CT，能谱CT在物质识别方面具有明显的优势，被认为是未来CT的发展方向之一。然而，作为核心器件的光子计数探测器目前还存在某些弊端：ASIC电路的复杂度限制了能区的数量，同时多能区设计导致每个能区数据信噪比下降。为克服上述难题，本项目组提出了一种动态双能区光子计数探测器的新设计思路，仅使用两个动态变化的高低能区完成整个能谱内光子的采集，其优势在于：通过CT扫描过程中动态调整每个探测器像素的能区阈值，理论上可以获得任意多个能区的数据，有可能获得比现有技术更高的能量分辨率、空间分辨率和更低的成本。本项目将针对动态双能区能谱CT，研究其X射线成像的物理机制和数学模型，以及能谱CT图像重建方法和材料分解方法，利用蒙卡模拟及现有实验平台对上述理论和方法进行验证。作为一种新的能谱CT设计理念，一旦本项目验证了其成像理论和方法的可行性，将会对本领域探测器及新型能谱CT的发展产生重要影响。

本项目针对动态双能区X射线能谱CT，研究了其成像的物理机制和数学模型，以及能谱CT图像重建方法和材料分解方法，并对现有实验平台进行了升级改造和实验验证。本项目创新性地提出了适用于动态双能区能谱CT图像重建的方法和材料分解方法，创新性地提出了基于去光电、动态双基材料和图像域聚类的能谱CT材料分解方法，创新地提出了X射线能谱和荧光CT双模态同时成像技术。利用蒙卡模拟和真实实验对上述理论和方法进行了验证和优化改进，为研制新型能谱CT成像系统，以及其在工业、安检、医疗等领域的应用推广奠定了理论和实验基础。其中高能双能CT重建和材料分解相关技术已经成功应用于世界首套大型集装箱高能双能CT产品中，开启了海关集装箱三维CT成像和智能审图的新时代。