Abstract:Inverse problems spanning four or more dimensions such as space, time andother independent parameters have become increasingly important. State-of-the-art 4D reconstruction methods use model based iterative reconstruction (MBIR), but depend critically on the quality of the prior modeling. Recently, plug-and-play (PnP) methods have been shown to be an effective way to incorporate advanced prior models using state-of-the-art denoising algorithms. However, state-of-the-art denoisers such as BM4D and deep convolutional neural networks (CNNs) are primarily available for 2D or 3D images and extending them to higher dimensions is difficult due to algorithmic complexity and the increased difficulty of effective training. In this paper, we present multi-slice fusion, a novel algorithm for 4D reconstruction, based on the fusion of multiple low-dimensional denoisers. Our approach uses multi-agent consensus equilibrium (MACE), an extension of plug-and-play, as a framework for integrating the multiple lower-dimensional models. We apply our method to 4D cone-beam X-ray CT reconstruction for non destructive evaluation (NDE) of samples that are dynamically moving during acquisition. We implement multi-slice fusion on distributed, heterogeneous clusters in order to reconstruct large 4D volumes in reasonable time and demonstrate the inherent parallelizable nature of the algorithm. We present simulated and real experimental results on sparse-view and limited-angle CT data to demonstrate that multi-slice fusion can substantially improve the quality of reconstructions relative to traditional methods, while also being practical to implement and train.

Multicomponent Volume Reconstruction from Slice Data Using a Modified Multicomponent Cahn–Hilliard System

Multi-slice Ptychographic Imaging with Multistage Coarse-to-fine Reconstruction

Three-dimensional Volume Reconstruction from Slice Data Using Phase-Field Models

Weighted 3D Volume Reconstruction from Series of Slice Data Using a Modified Allen–Cahn Equation

Volumetric Data Modeling and Analysis Based on Seven-Directional Box Spline

Irregular Object Measurement Method Based on Improved Adaptive Slicing Method

Efficient Boundary Surface Reconstruction from Heterogeneous Volumetric Data Via Tri-Prism Decomposition

Volume Reconstruction Based on Non-rigid Registration

Model-driven Multicomponent Volume Exploration.

Fast and Efficient Narrow Volume Reconstruction from Scattered Data.

Surface Carving-Based Automatic Volume Data Reduction

A novel segmentation algorithm for complex 3D mesh model in computer vision.

A New Method for Volume Construction of 3D Objects in 3D Images

Multiscale H-Dome Area Reconstruction Improved Transform With Applications To Image Segmentation

Dynamic Extraction For Voi From Ct Images Based On Volume Rendering

Model-based Multi-Component Volume Data Labeling and Rendering.

Multi-Slice Fusion for Sparse-View and Limited-Angle 4D CT Reconstruction

Three-dimensions volume reconstruction of grayscale serial slice images

Volume Measurement of Irregular Objects Based on Improved Point Cloud Slicing Method

Multiple Photo-Consistency Constraints for 3D Volumetric Reconstruction in Multi-View

A New Method for Contour Correspondence and Slice Registration Based on the Internal Information of Images