基于结构与动态联合先验的PET心肌灌注直接4D参数成像方法

Myocardial perfusion PET imaging, as an examination method for myocardial ischemia detection, has been the most important imaging method for diagnosis, evaluation, and prognosis of coronary heart disease. In particular, the absolutely quantification physiological parameters can be derived from the kinetic model, which have very important clinical significance. The present PET divides the image reconstruction and kinetic parameter estimation into two parts, and does not consider the statistical properties of detected data. Therefore, the error of reconstructed dynamic images will introduced into the kinetic parameter estimation, which results in lower signal to noise (SNR) of parametric image...The object of this proposal is to develop algorithms for structure and dynamic guide prior based direct 4D parametric imaging, mainly including: (1) Considering the kinetic model, the dynamic PET images was expressed by kinetic parameters, and then the direct 4D parametric image reconstruction was built from the detected data and kinetic parameters; (2) Based on the characteristic of MPI, the structure of prior image and dynamic of estimated parameters were combined to form a joint prior model, and incorporated in to the direct 4D parametric image reconstruction; (3) The analysis and evaluation metric of MPI was studied to optimize the parametric imaging model to improve the quality of reconstructed parametric image and provide accuracy and reliability evidence for the treatment of coronary heart disease.

PET心肌灌注显像(Myocardial perfusion imaging，MPI)作为一种检测心肌缺血的手段，已成为冠心病诊断、疗效评价及预后判断的重要影像学方法。特别是通过动力学模型的应用可以得到定量的生理学参数，具有重要的临床意义。当前PET将图像重建与参数估计分为两部分，动态PET重建图像误差被引入参数估计，导致参数图像极低的信噪比,极大的限制了其临床应用。鉴于此，本项目拟开展基于结构与动态联合先验的PET心肌灌注直接4D参数成像方法，主要包括:(1)从动力学模型出发，将动态PET图像参数化表示，利用探测数据统计特性，建立由探测数据到动力学参数的直接4D重建模型；(2)结合MPI扫描特点，构建基于先验图像结构与参数图像动态联合先验模型，引入直接4D重建模型；(3)开展PET心肌灌注显像参数图像分析与评价研究，优化参数成像模型，以期提高参数图像质量，为冠心病患者的治疗提供可靠依据。

PET心肌灌注显像(Myocardial perfusion imaging，MPI)作为一种检测心肌缺血的手段 ，已成为冠心病诊断、疗效评价及预后判断的重要影像学方法。特别是通过动力学模型的应用 可以得到定量的生理学参数，具有重要的临床意义。当前PET将图像重建与参数估计分为两部 分，动态PET重建图像误差被引入参数估计，导致参数图像极低的信噪比,极大的限制了其临床 应用。鉴于此，本项目拟开展基于结构与动态联合先验的PET心肌灌注直接4D参数成像方法， 主要包括:(1)从动力学模型出发，将动态PET图像参数化表示，利用探测数据统计特性，建立 由探测数据到动力学参数的直接4D重建模型；(2)结合MPI扫描特点，构建基于先验图像结构与 参数图像动态联合先验模型，引入直接4D重建模型；(3)开展PET心肌灌注显像参数图像分析与 评价研究，优化参数成像模型，以期提高参数图像质量，为冠心病患者的治疗提供可靠依据。项目取得较丰富的研究成果。项目合计发表期刊论文 28 篇，其中 SCI 论文 26 篇， 包括医学影像领域的主流期刊《European Radiology》、《Molecular Imaging and Biology》、《Computer Methods and Programs in Biomedicine》、《Medical Physics》，《Physics in Medicine and Biology》、《IEEE Journal of Biomedical and Health Informatics》等，中文核心期刊 2 篇。申请中国发明专利4项，其中3项获得发明授权，1项进入实质审查期。