DeepAI AI Chat
Log In Sign Up

Learned Primal-dual Reconstruction

by   Jonas Adler, et al.
KTH Royal Institute of Technology

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 PSNR 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.


page 5

page 7

page 8

page 9


Deep Learning Framework for Digital Breast Tomosynthesis Reconstruction

Digital breast tomosynthesis is rapidly replacing digital mammography as...

Learning-based Image Reconstruction via Parallel Proximal Algorithm

In the past decade, sparsity-driven regularization has led to advancemen...

Learned Digital Back-Propagation for Dual-Polarization Dispersion Managed Systems

Digital back-propagation (DBP) and learned DBP (LDBP) are proposed for n...

Computed Tomography Reconstruction Using Deep Image Prior and Learned Reconstruction Methods

In this work, we investigate the application of deep learning methods fo...