3DGR-CT: Sparse-View CT Reconstruction with a 3D Gaussian Representation

Abstract

Sparse-view computed tomography (CT) reduces radiation exposure by acquiring fewer projections, making it a valuable tool in clinical scenarios where low-dose radiation is essential. However, this often results in increased noise and artifacts due to limited data. In this paper we propose a novel 3D Gaussian representation (3DGR) based method for sparse-view CT reconstruction. Inspired by recent success in novel view synthesis driven by 3D Gaussian splatting, we leverage the efficiency and expressiveness of 3D Gaussian representation as an alternative to implicit neural representation. To unleash the potential of 3DGR for CT imaging scenario, we propose two key innovations: (i) FBP-image-guided Guassian initialization and (ii) efficient integration with a differentiable CT projector. Extensive experiments and ablations on diverse datasets demonstrate the proposed 3DGR-CT consistently outperforms state-of-the-art counterpart methods, achieving higher reconstruction accuracy with faster convergence. Furthermore, we showcase the potential of 3DGR-CT for real-time physical simulation, which holds important clinical applications while challenging for implicit neural representations.

References in corpus (22)

• Denoising Diffusion Probabilistic Models
• Instant Neural Graphics Primitives with a Multiresolution Hash Encoding
• Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
• VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-detector CT Images
• Deep-neural-network based sinogram synthesis for sparse-view CT image reconstruction
• Diffusion Posterior Sampling for General Noisy Inverse Problems
• Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models
• Solving Inverse Problems in Medical Imaging with Score-Based Generative Models
• SynthRAD2023 Grand Challenge dataset: generating synthetic CT for radiotherapy
• NAF: Neural Attenuation Fields for Sparse-View CBCT Reconstruction
• AdR-Gaussian: Accelerating Gaussian Splatting with Adaptive Radius
• FreeSeed: Frequency-band-aware and Self-guided Network for Sparse-view CT Reconstruction
• CoIL: Coordinate-based Internal Learning for Imaging Inverse Problems
• SNAF: Sparse-view CBCT Reconstruction with Neural Attenuation Fields
• 3D Teeth Reconstruction from Panoramic Radiographs using Neural Implicit Functions
• Solving Linear Inverse Problems Provably via Posterior Sampling with Latent Diffusion Models
• EndoGS: Deformable Endoscopic Tissues Reconstruction with Gaussian Splatting
• DDGS-CT: Direction-Disentangled Gaussian Splatting for Realistic Volume Rendering
• GaSpCT: Gaussian Splatting for Novel CT Projection View Synthesis
• R${}^2$-Gaussian: Rectifying Radiative Gaussian Splatting for Tomographic Reconstruction
• TOGS: Gaussian Splatting with Temporal Opacity Offset for Real-Time 4D DSA Rendering
• Computed tomography coronary angiogram images, annotations and associated data of normal and diseased arteries