Learned Primal-Dual Reconstruction
Learned Primal-Dual Reconstruction
复制标题
DOI：
10.1109/tmi.2018.2799231
复制
发表时间：
2018-06-01
影响因子：
10.6
通讯作者：
Oktem, Ozan
中科院分区：
文献类型：
--
作者：
Adler, Jonas;Oktem, Ozan
我们提出了用于断层扫描重建的学习原始对偶算法。该算法通过展开近端原始对偶优化方法来解释深度神经网络中的（可能是非线性的）前向算子，但其中近端算子已被卷积神经网络取代。该算法经过端到端训练，直接根据原始测量数据进行工作，并且不依赖于任何初始重建，例如滤波反投影 (FBP)。我们比较了所提出的方法在低剂量计算机断层扫描重建方面与 FBP、总变分 (TV) 和基于深度学习的 FBP 后处理的性能。对于 Shepp-Logan 模型，与所有比较方法相比，我们获得了 >6 dB 峰值信噪比改进。对于人类模型，相应的改进是比电视提高 6.6 dB，比学习后处理提高 2.2 dB，同时结构相似性指数也有显着提高。最后，我们的算法仅涉及十次正向-反向投影计算，使得该方法对于时间关键的临床应用是可行的。
We propose the Learned Primal-Dual algorithm for tomographic reconstruction. The algorithm accounts for a (possibly non-linear) forward operator in a deep neural network by unrolling a proximal primal-dual optimization method, but where the proximal operators have been replaced with convolutional neural networks. The algorithm is trained end-to-end, working directly from raw measured data and it does not depend on any initial reconstruction such as filtered back-projection (FBP). We compare performance of the proposed method on low dose computed tomography reconstruction against FBP, total variation (TV), and deep learning based post-processing of FBP. For the Shepp-Logan phantom we obtain >6 dB peak signal to noise ratio improvement against all compared methods. For human phantoms the corresponding improvement is 6.6 dB over TV and 2.2 dB over learned post-processing along with a substantial improvement in the structural similarity index. Finally, our algorithm involves only ten forward-back-projection computations, making the method feasible for time critical clinical applications.