Schleyer Paul J, O'Doherty Michael J, Barrington Sally F, Marsden Paul K
St Thomas' School of Medicine, King's College London, London, UK.
Phys Med Biol. 2009 Apr 7;54(7):1935-50. doi: 10.1088/0031-9155/54/7/005. Epub 2009 Mar 5.
Respiratory motion can adversely affect both PET and CT acquisitions. Respiratory gating allows an acquisition to be divided into a series of motion-reduced bins according to the respiratory signal, which is typically hardware acquired. In order that the effects of motion can potentially be corrected for, we have developed a novel, automatic, data-driven gating method which retrospectively derives the respiratory signal from the acquired PET and CT data. PET data are acquired in listmode and analysed in sinogram space, and CT data are acquired in cine mode and analysed in image space. Spectral analysis is used to identify regions within the CT and PET data which are subject to respiratory motion, and the variation of counts within these regions is used to estimate the respiratory signal. Amplitude binning is then used to create motion-reduced PET and CT frames. The method was demonstrated with four patient datasets acquired on a 4-slice PET/CT system. To assess the accuracy of the data-derived respiratory signal, a hardware-based signal was acquired for comparison. Data-driven gating was successfully performed on PET and CT datasets for all four patients. Gated images demonstrated respiratory motion throughout the bin sequences for all PET and CT series, and image analysis and direct comparison of the traces derived from the data-driven method with the hardware-acquired traces indicated accurate recovery of the respiratory signal.
呼吸运动可能会对PET和CT采集产生不利影响。呼吸门控允许根据呼吸信号将采集过程划分为一系列运动减少的时间段，呼吸信号通常由硬件采集。为了有可能校正运动的影响，我们开发了一种新颖的、自动的、数据驱动的门控方法，该方法从采集到的PET和CT数据中回顾性地导出呼吸信号。PET数据以列表模式采集并在正弦图空间中分析，CT数据以电影模式采集并在图像空间中分析。频谱分析用于识别CT和PET数据中受呼吸运动影响的区域，这些区域内计数的变化用于估计呼吸信号。然后使用幅度分箱来创建运动减少的PET和CT帧。该方法在使用4层PET/CT系统采集的四个患者数据集中得到了验证。为了评估数据衍生呼吸信号的准确性，采集了基于硬件的信号进行比较。对所有四名患者的PET和CT数据集成功进行了数据驱动的门控。门控图像显示了所有PET和CT系列在整个时间段序列中的呼吸运动，并且图像分析以及将数据驱动方法得到的轨迹与硬件采集的轨迹进行直接比较表明呼吸信号得到了准确恢复。
