Self-Supervised Slice-to-Volume Reconstruction with Gaussian Representations for Fetal MRI

2026-01-30

AI Summary

Fetal MRI is frequently degraded by motion artifacts, and conventional 3D reconstruction methods rely on multiple orthogonal stacks and are computationally intensive. Moreover, existing learning-based approaches require ground-truth 3D labels, which are difficult to obtain in clinical practice. To address these challenges, this work proposes GaussianSVR, a novel self-supervised framework that introduces 3D Gaussian representations into fetal MRI reconstruction for the first time. The method generates self-supervision by simulating the forward slice-acquisition process and jointly optimizes Gaussian parameters alongside spatial transformations. A multi-resolution strategy is employed to enhance both reconstruction accuracy and computational efficiency. Notably, GaussianSVR eliminates the need for real 3D annotations and achieves significant improvements over current baselines in both image quality and speed.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: 3D Computer Vision
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Reconstructing 3D fetal MR volumes from motion-corrupted stacks of 2D slices is a crucial and challenging task. Conventional slice-to-volume reconstruction (SVR) methods are time-consuming and require multiple orthogonal stacks for reconstruction. While learning-based SVR approaches have significantly reduced the time required at the inference stage, they heavily rely on ground truth information for training, which is inaccessible in practice. To address these challenges, we propose GaussianSVR, a self-supervised framework for slice-to-volume reconstruction. GaussianSVR represents the target volume using 3D Gaussian representations to achieve high-fidelity reconstruction. It leverages a simulated forward slice acquisition model to enable self-supervised training, alleviating the need for ground-truth volumes. Furthermore, to enhance both accuracy and efficiency, we introduce a multi-resolution training strategy that jointly optimizes Gaussian parameters and spatial transformations across different resolution levels. Experiments show that GaussianSVR outperforms the baseline methods on fetal MR volumetric reconstruction. Code will be available upon acceptance.

Problem

• fetal MRI
• slice-to-volume reconstruction
• motion corruption
• self-supervised learning
• 3D reconstruction

Innovation

• self-supervised learning
• Gaussian representations
• slice-to-volume reconstruction
• fetal MRI
• multi-resolution optimization