Fu Jintao, Cong Peng, Xu Shuo, Chang Jiahao, Liu Ximing, Sun Yuewen
Institute of Nuclear and New Energy Technology, Tsinghua University, Beijing, China.
Beijing Key Laboratory of Nuclear Detection Technology, Beijing, China.
Med Phys. 2025 May;52(5):3044-3058. doi: 10.1002/mp.17685. Epub 2025 Feb 10.
Sparse-view computed tomography (CT) reduces radiation exposure but suffers from severe artifacts caused by insufficient sampling and data scarcity, which compromise image fidelity. Recent advancements in deep learning (DL)-based methods for inverse problems have shown promise for CT reconstruction but often require high-quality paired datasets and lack interpretability.
This paper aims to advance the field of CT reconstruction by introducing a novel unsupervised deep learning method. It builds on the foundation of Deep Radon Prior (DRP), which utilizes an untrained encoder-decoder network to extract implicit features from the Radon domain, and leverages Neural Architecture Search (NAS) to optimize network structures.
We propose a novel unsupervised deep learning method for image reconstruction, termed NAS-DRP. This method leverages reinforcement learning-based NAS to explore diverse architectural spaces and integrates reinforcement learning with data inconsistency in the Radon domain. Building on previous DRP research, NAS-DRP utilizes an untrained encoder-decoder network to extract implicit features from the Radon domain. It further incorporates insights from studies on Deep Image Prior (DIP) regarding the critical impact of upsampling layers on image quality restoration. The method employs NAS to search for the optimal network architecture for upsampling unit tasks, while using Recurrent Neural Networks (RNNs) to constrain the optimization process, ensuring task-specific improvements in sparse-view CT image reconstruction.
Extensive experiments demonstrate that the NAS-DRP method achieves significant performance improvements in multiple CT image reconstruction tasks. The proposed method outperforms traditional reconstruction methods and other DL-based techniques in terms of both objective metrics (PSNR, SSIM, and LPIPS) and subjective visual quality. By automatically optimizing network structures, NAS-DRP effectively enhances the detail and accuracy of reconstructed images while minimizing artifacts.
NAS-DRP represents a significant advancement in the field of CT image reconstruction. By integrating NAS with deep learning and leveraging Radon domain-specific adaptations, this method effectively addresses the inherent challenges of sparse-view CT imaging. Additionally, it reduces the cost and complexity of data acquisition, demonstrating substantial potential for broader application in medical imaging. The evaluation code will be available at https://github.com/fujintao1999/NAS-DRP/.
稀疏视图计算机断层扫描（CT）可减少辐射暴露，但会因采样不足和数据稀缺而产生严重伪影，这会损害图像保真度。基于深度学习（DL）的反问题方法的最新进展已显示出在CT重建方面的前景，但通常需要高质量的配对数据集且缺乏可解释性。
本文旨在通过引入一种新颖的无监督深度学习方法来推动CT重建领域的发展。它建立在深度拉东先验（DRP）的基础上，该方法利用未训练的编码器-解码器网络从拉东域中提取隐含特征，并利用神经架构搜索（NAS）来优化网络结构。
我们提出了一种用于图像重建的新颖无监督深度学习方法，称为NAS-DRP。该方法利用基于强化学习的NAS来探索不同的架构空间，并将强化学习与拉东域中的数据不一致性相结合。基于先前的DRP研究，NAS-DRP利用未训练的编码器-解码器网络从拉东域中提取隐含特征。它进一步纳入了关于深度图像先验（DIP）的研究中有关上采样层对图像质量恢复的关键影响的见解。该方法采用NAS来搜索上采样单元任务的最佳网络架构，同时使用递归神经网络（RNN）来约束优化过程，确保在稀疏视图CT图像重建中针对特定任务的改进。
大量实验表明，NAS-DRP方法在多个CT图像重建任务中取得了显著的性能提升。所提出的方法在客观指标（PSNR、SSIM和LPIPS）和主观视觉质量方面均优于传统重建方法和其他基于DL的技术。通过自动优化网络结构，NAS-DRP有效地增强了重建图像的细节和准确性，同时将伪影降至最低。
NAS-DRP代表了CT图像重建领域的一项重大进展。通过将NAS与深度学习相结合并利用特定于拉东域的适应性，该方法有效地解决了稀疏视图CT成像的固有挑战。此外，它降低了数据采集的成本和复杂性，显示出在医学成像中更广泛应用的巨大潜力。评估代码将在https://github.com/f
