[论文解读] Deep-Fill: Deep Learning Based Sinogram Domain Gap Filling in Positron Emission Tomography

Isaac Shiri, Peyman Sheikhzadeh+1 more|arXiv (Cornell University)|Jun 16, 2019

Medical Imaging Techniques and Applications被引用 19

一句话总结

本文提出 Deep-Fill，一种深度卷积编码器-解码器网络，用于重建因探测器间间隙导致的 PET sinogram 中的缺失数据。通过在增强的幻影数据上进行训练，该模型能够高精度地预测无间隙的 sinogram，显著减少伪影并提升重建 PET 图像的质量。

ABSTRACT

One of the major challenges in design and developing of PET, scanners are the presence of inactive areas between the detector blocks which degrade the image spatial resolution and leads to streaking artifacts especially when we employ analytical image reconstruction. The aim of this study is to assess the feasibility of generating the gap-free PET image using the deep convolutional encoder-decoder in sinogram space. The gap-corrupted sinograms of simulated HRRT scanner, sinograms without gaps as ideal/ground truth and predicted sinograms owing to the implemented our deep-fill method were quantitatively compared. In total, 1293 phantom images divided into three main sets of training 1000, 150 Validation, and 143 test set. The 1000 training images were augmented using affine transformations with various sub-transforms including rotation (rotate randomly), translation to 12000 Image with 6 frequencies of 2 main methods. The Deep-Fill architecture consists of an encoder and a decoder part, and it is composed of convolution operation, max pooling, ReLU activation, concatenation, and up convolution layers. The gap image is going through the network along with all possible paths then the gap-free image was generated by decoder part of the network. The quality of the generated images was quantitatively assessed by different quality metrics in both sinogram space and reconstruction images. We demonstrated that deep learning based approaches applied to inter-detector gap filling can recover the missing data in sinogram with high quantitative accuracy and have the potential to significantly improve the reconstructed image quality and prevent degradation of PET image quantification.

研究动机与目标

• 解决探测器模块之间存在非活动区域所导致的 PET 图像质量下降问题。
• 探究深度学习是否能有效重建 sinogram 空间中的缺失数据。
• 开发一种在最小化条状伪影的同时保持图像定量精度的方法。
• 评估卷积自编码器在从受损输入生成无间隙 sinogram 方面的性能。

提出的方法

• 采用类似 U-Net 的编码器-解码器架构，包含卷积层、池化层、ReLU 激活函数、跳跃连接和转置卷积层。
• 网络在 12,000 张通过仿射变换从 1,000 个原始幻影研究中生成的增强训练图像上进行训练。
• 输入的 sinogram 包含模拟探测器间死区的的人工间隙；模型预测对应的无间隙 sinogram。
• 训练过程中，以无间隙的真实 sinogram 作为监督学习的目标。
• 数据增强包括随机旋转、平移以及多种频率子变换，以提升泛化能力。
• 通过 sinogram 空间和重建图像空间中的定量指标评估模型性能。

实验结果

研究问题

• RQ1 深度学习能否有效恢复因探测器间间隙导致的 PET sinogram 中的缺失数据？
• RQ2 卷积自编码器从受损输入重建无间隙 sinogram 的精度如何？
• RQ3 填补空隙在多大程度上提升了重建 PET 图像的质量？
• RQ4 与传统重建方法相比，所提出方法是否保持了图像的定量精度？

主要发现

• Deep-Fill 模型在重建无间隙 sinogram 方面实现了高定量精度，显著减少了条状伪影。
• 在重建图像中，图像质量指标在采用深度学习方法填补空隙后得到显著提升。
• 由于广泛的数据增强，该模型在多种幻影配置下均表现出强健的泛化能力。
• 该方法保持了图像的定量精度，表明其在临床中改善 PET 图像重建的可行性。