Supervised Deep Learning for Head Motion Correction in PET.
Supervised Deep Learning for Head Motion Correction in PET.
复制标题
PET 中头部运动校正的监督深度学习。
DOI：
10.1007/978-3-031-16440-8_19
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Onofrey,JohnA
中科院分区：
文献类型：
--
作者：
Zeng,Tianyi;Zhang,Jiazhen;Revilla,Enette;Lieffrig,EléonoreV;Fang,Xi;Lu,Yihuan;Onofrey,JohnA
头部运动是正电子发射断层扫描(PET)成像的一个主要限制因素，会导致图像伪影和量化误差。头部运动校正在神经系统疾病的定量图像分析和诊断中起着至关重要的作用。然而，到目前为止，还没有一种方法可以在不使用外部设备的情况下连续跟踪头部运动。在这里，我们开发了一个基于深度学习的算法，通过利用现有的动态PET扫描和来自外部Polaris Vicra跟踪的黄金标准运动测量来预测脑部PET的刚性运动。我们提出了一种用于头部运动校正的深度学习(DL-HMC)方法，该方法由三个部分组成：(I)PET输入数据编码层；(Ii)用于估计六个刚体运动变换参数的回归层；(Iii)用于调节网络以跟踪时间活动的基于特征的变换(FWT)层。DL-HMC的输入是对PET数据的1秒3D云表示的采样对，输出是对6个刚体变换运动参数的预测。我们使用Vicra运动跟踪信息作为金标准，以有监督的方式训练这个网络。我们通过与金标准维克拉测量进行比较来定量评估DL-HMC，定性评估重建图像以及执行感兴趣区标准摄取值(SUV)测量。我们进行了一项算法烧蚀研究，以确定我们的每个DL-HMC设计选择对网络性能的贡献。我们的结果表明，在没有外部运动跟踪硬件的情况下，使用数据驱动的配准方法可以对脑PET进行准确的运动预测。所有代码都可在giHub上公开获得：https://github.com/OnofreyLab/dl-hmc_miccai2022.
Head movement is a major limitation in brain positron emission tomography (PET) imaging, which results in image artifacts and quantification errors. Head motion correction plays a critical role in quantitative image analysis and diagnosis of nervous system diseases. However, to date, there is no approach that can track head motion continuously without using an external device. Here, we develop a deep learning-based algorithm to predict rigid motion for brain PET by leveraging existing dynamic PET scans with gold-standard motion measurements from external Polaris Vicra tracking. We propose a novel Deep Learning for Head Motion Correction (DL-HMC) methodology that consists of three components: (i) PET input data encoder layers; (ii) regression layers to estimate the six rigid motion transformation parameters; and (iii) feature-wise transformation (FWT) layers to condition the network to tracer time-activity. The input of DL-HMC is sampled pairs of one-second 3D cloud representations of the PET data and the output is the prediction of six rigid transformation motion parameters. We trained this network in a supervised manner using the Vicra motion tracking information as gold-standard. We quantitatively evaluate DL-HMC by comparing to gold-standard Vicra measurements and qualitatively evaluate the reconstructed images as well as perform region of interest standard uptake value (SUV) measurements. An algorithm ablation study was performed to determine the contributions of each of our DL-HMC design choices to network performance. Our results demonstrate accurate motion prediction performance for brain PET using a data-driven registration approach without external motion tracking hardware. All code is publicly available on GitHub: https://github.com/OnofreyLab/dl-hmc_miccai2022.