GR-Gaussian: Graph-Based Radiative Gaussian Splatting for Sparse-View CT Reconstruction

TL;DR

GR-Gaussian introduces a graph-based 3D Gaussian Splatting framework that effectively reduces artifacts and enhances accuracy in sparse-view CT reconstruction by refining gradient computation and improving initialization.

Contribution

It presents a novel graph-based approach with a denoised initialization and pixel-graph-aware gradient strategy to improve sparse-view CT reconstruction accuracy.

Findings

Achieves PSNR improvements of 0.67 dB and 0.92 dB.

Attains SSIM gains of 0.011 and 0.021.

Effectively suppresses needle-like artifacts in sparse-view conditions.

Abstract

3D Gaussian Splatting (3DGS) has emerged as a promising approach for CT reconstruction. However, existing methods rely on the average gradient magnitude of points within the view, often leading to severe needle-like artifacts under sparse-view conditions. To address this challenge, we propose GR-Gaussian, a graph-based 3D Gaussian Splatting framework that suppresses needle-like artifacts and improves reconstruction accuracy under sparse-view conditions. Our framework introduces two key innovations: (1) a Denoised Point Cloud Initialization Strategy that reduces initialization errors and accelerates convergence; and (2) a Pixel-Graph-Aware Gradient Strategy that refines gradient computation using graph-based density differences, improving splitting accuracy and density representation. Experiments on X-3D and real-world datasets validate the effectiveness of GR-Gaussian, achieving PSNR improvements of 0.67 dB and 0.92 dB, and SSIM gains of 0.011 and 0.021. These results highlight the applicability of GR-Gaussian for accurate CT reconstruction under challenging sparse-view conditions.