MoRe-3DGSMR: Motion-resolved reconstruction framework for free-breathing pulmonary MRI based on 3D Gaussian representation

2025-05-08

AI Summary

Severe motion artifacts in free-breathing lung MRI hinder high-resolution, motion-resolved reconstruction. Method: We propose an unsupervised motion-resolved 3D isotropic reconstruction framework. It integrates golden-angle radial sampling with k-space center-based respiratory signal estimation to acquire multi-phase data. We introduce 3D Gaussian Splatting (3DGS) into lung MRI reconstruction to model a continuous, smooth reference volume and design a patient-specific CNN to estimate registration-free ground-truth deformation fields. Multi-phase images are synthesized via spatial transformation. Results: Evaluated on six subjects, our method significantly outperforms three state-of-the-art approaches, achieving substantial improvements in SNR and CNR. It is the first to produce high-quality, motion-artifact-free dynamic lung MR images at 1 mm isotropic resolution—enabling unprecedented visualization of pulmonary dynamics without breath-holding or external motion monitoring.

Technology Category

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

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

This study presents an unsupervised, motion-resolved reconstruction framework for high-resolution, free-breathing pulmonary magnetic resonance imaging (MRI), utilizing a three-dimensional Gaussian representation (3DGS). The proposed method leverages 3DGS to address the challenges of motion-resolved 3D isotropic pulmonary MRI reconstruction by enabling data smoothing between voxels for continuous spatial representation. Pulmonary MRI data acquisition is performed using a golden-angle radial sampling trajectory, with respiratory motion signals extracted from the center of k-space in each radial spoke. Based on the estimated motion signal, the k-space data is sorted into multiple respiratory phases. A 3DGS framework is then applied to reconstruct a reference image volume from the first motion state. Subsequently, a patient-specific convolutional neural network is trained to estimate the deformation vector fields (DVFs), which are used to generate the remaining motion states through spatial transformation of the reference volume. The proposed reconstruction pipeline is evaluated on six datasets from six subjects and bench-marked against three state-of-the-art reconstruction methods. The experimental findings demonstrate that the proposed reconstruction framework effectively reconstructs high-resolution, motion-resolved pulmonary MR images. Compared with existing approaches, it achieves superior image quality, reflected by higher signal-to-noise ratio and contrast-to-noise ratio. The proposed unsupervised 3DGS-based reconstruction method enables accurate motion-resolved pulmonary MRI with isotropic spatial resolution. Its superior performance in image quality metrics over state-of-the-art methods highlights its potential as a robust solution for clinical pulmonary MR imaging.

Problem

• Reconstructs high-resolution pulmonary MRI during free-breathing
• Addresses motion artifacts using 3D Gaussian representation
• Improves image quality with unsupervised neural network deformation

Innovation

• Uses 3D Gaussian representation for motion-resolved MRI
• Employs golden-angle radial sampling for data acquisition
• Trains CNN to estimate deformation vector fields