Division of Nuclear Medicine and Molecular Imaging, Department of Medical Imaging, Geneva University Hospital, Geneva, Switzerland.
Department of Psychiatry, Geneva University, Geneva, Switzerland.
Hum Brain Mapp. 2020 Sep;41(13):3667-3679. doi: 10.1002/hbm.25039. Epub 2020 May 21.
PET attenuation correction (AC) on systems lacking CT/transmission scanning, such as dedicated brain PET scanners and hybrid PET/MRI, is challenging. Direct AC in image-space, wherein PET images corrected for attenuation and scatter are synthesized from nonattenuation corrected PET (PET-nonAC) images in an end-to-end fashion using deep learning approaches (DLAC) is evaluated for various radiotracers used in molecular neuroimaging studies. One hundred eighty brain PET scans acquired using F-FDG, F-DOPA, F-Flortaucipir (targeting tau pathology), and F-Flutemetamol (targeting amyloid pathology) radiotracers (40 + 5, training/validation + external test, subjects for each radiotracer) were included. The PET data were reconstructed using CT-based AC (CTAC) to generate reference PET-CTAC and without AC to produce PET-nonAC images. A deep convolutional neural network was trained to generate PET attenuation corrected images (PET-DLAC) from PET-nonAC. The quantitative accuracy of this approach was investigated separately for each radiotracer considering the values obtained from PET-CTAC images as reference. A segmented AC map (PET-SegAC) containing soft-tissue and background air was also included in the evaluation. Quantitative analysis of PET images demonstrated superior performance of the DLAC approach compared to SegAC technique for all tracers. Despite the relatively low quantitative bias observed when using the DLAC approach, this approach appears vulnerable to outliers, resulting in noticeable local pseudo uptake and false cold regions. Direct AC in image-space using deep learning demonstrated quantitatively acceptable performance with less than 9% absolute SUV bias for the four different investigated neuroimaging radiotracers. However, this approach is vulnerable to outliers which result in large local quantitative bias.
在缺乏 CT/透射扫描的系统（如专用脑部 PET 扫描仪和 PET/MRI 混合系统）上进行 PET 衰减校正（AC）具有挑战性。在图像空间中直接进行 AC，使用深度学习方法（DLAC），从非衰减校正的 PET（PET-nonAC）图像端到端合成经过衰减和散射校正的 PET 图像，用于评估各种在分子神经影像学研究中使用的放射性示踪剂。共纳入了 180 例使用 F-FDG、F-DOPA、F-Flortaucipir（靶向 tau 病理学）和 F-Flutemetamol（靶向淀粉样蛋白病理学）放射性示踪剂进行的脑部 PET 扫描（40+5，训练/验证+外部测试，每种示踪剂的受试者）。使用基于 CT 的 AC（CTAC）重建 PET 数据以生成参考 PET-CTAC，并在没有 AC 的情况下生成 PET-nonAC 图像。训练了一个深度卷积神经网络，从 PET-nonAC 生成 PET 衰减校正图像（PET-DLAC）。分别考虑从 PET-CTAC 图像获得的值作为参考，针对每种示踪剂研究了这种方法的定量准确性。还在评估中包括包含软组织和背景空气的分段 AC 图（PET-SegAC）。与 SegAC 技术相比，PET 图像的定量分析显示，对于所有示踪剂，DLAC 方法的性能均优于后者。尽管使用 DLAC 方法时观察到相对较低的定量偏差，但该方法容易受到异常值的影响，导致明显的局部假性摄取和假性冷区。使用深度学习在图像空间中进行直接 AC，对于四种不同的神经影像学放射性示踪剂，其定量偏差小于 9%，表现出可接受的性能。但是，这种方法容易受到异常值的影响，从而导致较大的局部定量偏差。
