UniSpine-GS: An Efficient Physics-Aware Gaussian Framework for Cross-Modality Multi-view Spine Image Synthesis

2026-07-06

AI Summary

Current 3D spinal assessment is hindered by the high cost of imaging equipment and physical discrepancies across modalities, making it difficult to achieve both accuracy and generalizability. This work proposes a physics-aware Gaussian framework that bypasses explicit 3D reconstruction, leveraging geometrically informed Gaussian representations to synthesize consistent multi-view spinal images. It further introduces a novel structure-preserving loss reweighting mechanism (SPWM) that substantially enhances anatomical boundary delineation and structural fidelity. As the first study to apply physics-aware Gaussian representations to cross-modal, multi-view spinal image synthesis, the proposed method significantly outperforms existing approaches on both the CTSpine3D benchmark and the newly curated FeSpine3D dataset, offering an efficient, low-cost, and anatomically consistent unified solution for medical imaging.

Abstract

The diagnosis of spinal diseases is often assisted by 3D imaging techniques in clinical practice. However, precise 3D spinal assessment is limited by the high costs of 3D imaging hardware and the challenges posed by the physical differences between imaging modalities, which hinder the generalizability of models. To address these issues, we propose UniSpine-GS, an efficient, physics-aware Gaussian framework designed for novel-view projection rendering in multi-view spine imaging via a 3D-aware representation. Instead of performing explicit 3D reconstruction, our approach learns a geometry-aware Gaussian representation that ensures anatomical consistency across different views. We introduce SPWM, a structure-guided loss reweighting strategy to improve boundary fidelity and local details. We evaluate our method on the CTSpine3D dataset and a newly constructed 3D fetal ultrasound dataset, FeSpine3D. Our results demonstrate that UniSpine-GS significantly outperforms existing methods across all metrics, offering a practical and cost-effective solution for unified multi-view medical imaging. Our code is publicly available at https://github.com/orangeisland66/UniSpine-GS.

Problem

• spine imaging
• cross-modality
• multi-view synthesis
• 3D medical imaging
• image generalizability

Innovation

• Gaussian representation
• physics-aware modeling
• multi-view synthesis
• structure-guided loss
• cross-modality imaging