ADAPTIVE-NET: deep computed tomography reconstruction network with analytical domain transformation knowledge
ADAPTIVE-NET: deep computed tomography reconstruction network with analytical domain transformation knowledge
复制标题
DOI：
10.21037/qims.2019.12.12
复制
发表时间：
2020-02-01
影响因子：
2.8
通讯作者：
Liang, Dong
中科院分区：
文献类型：
--
作者：
Ge, Yongshuai;Su, Ting;Liang, Dong
背景：最近，随着深度学习技术的发展，计算机断层扫描（CT）重建的范式发生了变化。在本研究中，我们提出了一种新的卷积神经网络（称为 ADAPTIVE-NET），通过集成分析域变换知识，直接从正弦图执行 CT 图像重建。方法：在所提出的 ADAPTIVE-NET 中，定制了具有恒定权重的特定网络层，通过分析反投影将正弦图变换到 CT 图像域。通过这个新框架，可以在正弦图域和 CT 图像域上同时执行特征提取。 Mayo 低剂量 CT (LDCT) 数据用于验证新网络。特别是，新网络与之前提出的残差编码器-解码器（RED）-CNN 网络进行了比较。对于每个网络，比较有和没有基于 VGG 的感知损失的均方误差 (MSE) 损失。此外，为了使用某些指标评估图像质量，通过每种方法重建的 LDCT 上的噪声功率谱 (NPS) 来量化噪声相关性。结果：可以通过 ADAPTIVE-NET 在具有中等内存大小（例如 11 GB）的单个图形处理单元 (GPU) 上轻松地从正弦图重建临床相关尺寸为 512x512 的图像。使用相同的 MSE 损失函数，新网络能够产生比 RED-CNN 更好的结果。此外，如果联合使用 VGG 损失，新网络能够重建自然的 CT 图像，并提高图像质量。结论：新提出的端到端监督 ADAPTIVE-NET 能够直接从正弦图重建高质量的 LDCT 图像。
Background: Recently, the paradigm of computed tomography (CT) reconstruction has shifted as the deep learning technique evolves. In this study, we proposed a new convolutional neural network (called ADAPTIVE-NET) to perform CT image reconstruction directly from a sinogram by integrating the analytical domain transformation knowledge.Methods: In the proposed ADAPTIVE-NET, a specific network layer with constant weights was customized to transform the sinogram onto the CT image domain via analytical back-projection. With this new framework, feature extractions were performed simultaneously on both the sinogram domain and the CT image domain. The Mayo low dose CT (LDCT) data was used to validate the new network. In particular, the new network was compared with the previously proposed residual encoder-decoder (RED)-CNN network. For each network, the mean square error (MSE) loss with and without VGG-based perceptual loss was compared. Furthermore, to evaluate the image quality with certain metrics, the noise correlation was quantified via the noise power spectrum (NPS) on the reconstructed LDCT for each method.Results: images that have clinically relevant dimensions of 512x512 can be easily reconstructed from a sinogram on a single graphics processing unit (GPU) with moderate memory size (e.g., 11 GB) by ADAPTIVE-NET. With the same MSE loss function, the new network is able to generate better results than the RED-CNN. Moreover, the new network is able to reconstruct natural looking CT images with enhanced image quality if jointly using the VGG loss.Conclusions: The newly proposed end-to-end supervised ADAPTIVE-NET is able to reconstruct high-quality LDCT images directly from a sinogram.