Swap-Net: A Memory-Efficient 2.5D Network for Sparse-View 3D Cone Beam CT Reconstruction

TL;DR

Swap-Net is a novel memory-efficient 2.5D deep learning architecture that effectively reconstructs 3D CBCT images from sparse projections, outperforming traditional and existing deep learning methods in quality and detail preservation.

Contribution

The paper introduces Swap-Net, a 2.5D network utilizing axes-swapping operations to enable high-quality 3D reconstruction without full 3D convolutions, addressing memory limitations.

Findings

Outperforms baseline methods in artifact reduction

Preserves details in complex hydrodynamic simulations

Demonstrates effectiveness in sparse-view 3D CBCT reconstruction

Abstract

Reconstructing 3D cone beam computed tomography (CBCT) images from a limited set of projections is an important inverse problem in many imaging applications from medicine to inertial confinement fusion (ICF). The performance of traditional methods such as filtered back projection (FBP) and model-based regularization is sub-optimal when the number of available projections is limited. In the past decade, deep learning (DL) has gained great popularity for solving CT inverse problems. A typical DL-based method for CBCT image reconstruction is to learn an end-to-end mapping by training a 2D or 3D network. However, 2D networks fail to fully use global information. While 3D networks are desirable, they become impractical as image sizes increase because of the high memory cost. This paper proposes Swap-Net, a memory-efficient 2.5D network for sparse-view 3D CBCT image reconstruction. Swap-Net uses a sequence of novel axes-swapping operations to produce 3D volume reconstruction in an end-to-end fashion without using full 3D convolutions. Simulation results show that Swap-Net consistently outperforms baseline methods both quantitatively and qualitatively in terms of reducing artifacts and preserving details of complex hydrodynamic simulations of relevance to the ICF community.