Multivariate Gaussian NeRF for Wide Field-of-View Ultrasound Reconstruction

2026-04-27

AI Summary

Traditional wide-field-of-view ultrasound imaging suffers from aliasing artifacts during stitching of convex-array probe scans due to depth-dependent resolution variations. This work proposes Ultra-Wide-NeRF, the first method to introduce a multivariate 3D Gaussian neural radiance field to this task. By employing distance-aware conical volume sampling and explicitly modeling the ultrasound beam geometry with anisotropic 3D Gaussians, Ultra-Wide-NeRF achieves continuous, aliasing-resistant neural tissue representation. The approach intrinsically suppresses stitching artifacts and enables high-quality rendering from arbitrary viewpoints. Validated on both phantom and in vivo porcine intracardiac ultrasound data, the method significantly extends the spatial context available for intraoperative navigation.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Wide Field-of-View (WFoV) reconstruction enhances 3D ultrasound imaging by providing valuable anatomical context for segmentation models and visualization. Clinical ultrasound volumes are predominantly acquired using convex probes, which generate expanding, diverging acoustic beams to maximize anatomical coverage. Stitching these sweeps together traditionally introduces significant compounding artifacts and aliasing due to depth-dependent resolution changes. Here, we introduce Ultra-Wide-NeRF, a Multivariate 3D Gaussian (MVG) NeRF-based method for WFoV ultrasound reconstruction. By explicitly modeling the complex beam geometry using distance-dependent convex volumetric sampling and anisotropic 3D Gaussians, our method inherently mitigates these compounding artifacts and provides anti-aliasing. Beyond simply reconstructing a static 3D grid, our NeRF-based approach yields a continuous neural representation of the tissue, enabling the synthesis of high-fidelity novel views from arbitrary virtual trajectories. We validate Ultra-Wide-NeRF for intracardiac echocardiography on phantom and porcine datasets, demonstrating that our method expands the spatial context important in intraoperative navigation. Code will be open-sourced upon publication.

Problem

• Wide Field-of-View
• ultrasound reconstruction
• compounding artifacts
• aliasing
• convex probe

Innovation

• Multivariate Gaussian NeRF
• Wide Field-of-View Ultrasound
• Anisotropic 3D Gaussians
• Neural Radiance Fields
• Ultrasound Reconstruction