College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou, China.
College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou, China; Zhejiang Provincial Key Laboratory of Information Processing, Communication and Networking, Hangzhou, China.
Magn Reson Imaging. 2020 Jul;70:134-144. doi: 10.1016/j.mri.2020.04.016. Epub 2020 Apr 28.
With the aim of developing a fast algorithm for high-quality MRI reconstruction from undersampled k-space data, we propose a novel deep neural Network, which is inspired by Iterative Shrinkage Thresholding Algorithm with Data consistency (NISTAD). NISTAD consists of three consecutive blocks: an encoding block, which models the flow graph of ISTA, a classical iteration-based compressed sensing magnetic resonance imaging (CS-MRI) method; a decoding block, which recovers the image from sparse representation; a data consistency block, which adaptively enforces consistency with the acquired raw data according to learned noise level. The ISTA method is thereby mapped to an end-to-end deep neural network, which greatly reduces the reconstruction time and simplifies the tuning of hyper-parameters, compared to conventional model-based CS-MRI methods. On the other hand, compared to general deep learning-based MRI reconstruction methods, the proposed method uses a simpler network architecture with fewer parameters. NISTAD has been validated on retrospectively undersampled diencephalon standard challenge data using different acceleration factors, and compared with DAGAN and Cascade CNN, two state-of-the-art deep neural network-based methods which outperformed many other state-of-the-art model-based and deep learning-based methods. Experimental results demonstrated that NISTAD reconstruction achieved comparable image quality with DAGAN and Cascade CNN reconstruction in terms of both PSNR and SSIM metrics, and subjective assessment, though with a simpler network structure.
为了开发一种从欠采样 k 空间数据中进行高质量 MRI 重建的快速算法，我们提出了一种新的深度神经网络，该网络受到基于迭代收缩阈值算法和数据一致性（NISTAD）的启发。NISTAD 由三个连续的模块组成：一个编码模块，它对 ISTA 的流程图进行建模，这是一种基于经典迭代的压缩感知磁共振成像（CS-MRI）方法；一个解码模块，它从稀疏表示中恢复图像；一个数据一致性模块，它根据学习到的噪声水平自适应地对原始数据进行一致性约束。这样，ISTA 方法就被映射到一个端到端的深度神经网络中，与传统的基于模型的 CS-MRI 方法相比，大大减少了重建时间，简化了超参数的调整。另一方面，与一般的基于深度学习的 MRI 重建方法相比，所提出的方法使用了更简单的网络结构和更少的参数。NISTAD 已经在不同加速因子的回顾性采样间脑标准挑战数据上进行了验证，并与 DAGAN 和级联 CNN 进行了比较，这两种最先进的基于深度学习的方法优于许多其他最先进的基于模型和基于深度学习的方法。实验结果表明，NISTAD 重建在 PSNR 和 SSIM 度量以及主观评估方面与 DAGAN 和级联 CNN 重建具有可比的图像质量，尽管网络结构更简单。
