Abstract: The degradation of image quality of PET/CT caused by respiratory motion will affect the physician’s diagnosis. The common developed technology of respiratory motion correction is gating, but it still has some limitations. This paper proposes a new method that using CT images extract the features of respiratory motion based on B-Spline to correct respiration. Firstly, it obtains the sequence of CT images corresponding with PET images within same respiratory motion cycle, and extracts the features of the motion of CT sequences within respiratory cycle by B-Spline. Next it transforms the feature parameters of CT sequences into the corresponding PET image sequences for the motion correction. The geometric deformation phantom and voxelised phantom tests show that the proposed method can obviously improve the quality of image of PET/CT for respiratory motion, and has a value of study further.

Key words: PET/CT imaging, respiratory motion correction, B-spline

摘要： PET/CT成像中的人体呼吸运动会造成图像运动模糊，会严重影响图像质量，对医生的诊断造成影响。目前常用的呼吸门控技术能够在一定程度上改善图像质量，但是均存在其局限性。提出了一种基于CT图像提取呼吸运动特征的B样条方法，对呼吸运动图像进行校正。在PET/CT上获取与PET图像周期匹配的CT图像序列，通过B样条配准方式对在呼吸周期内的CT图序列像提取运动信息；对CT图像所匹配的PET图像做基于运动特征信息的校正变换进行运动校正。几何位移形变和像素人体模实验结果表明，提出的方法对PET/CT呼吸运动图像质量改进明显，具有研究价值。

关键词: PET/CT成像, 呼吸运动校正, B样条