Primitive Representation Learning for Unsupervised Dynamic Contrast-Enhanced MRI Reconstruction

Forum topic · 小凯 · 2026-08-20

Summary

This paper introduces a multi-dimensional, primitive-based framework for unsupervised reconstruction of dynamic contrast-enhanced (DCE) MRI. Building on prior scan-specific methods that use Gaussian and Gabor primitives without large training datasets, the approach adds the dynamic contrast dimension. The framework disentangles the underlying anatomy, dynamic contrast enhancement, and residual motion into separate temporal basis functions, enabling a geometric interpretation of the learned representation. Experiments show the architecture achieves reconstruction quality and accuracy of extracted aortic and renal enhancement curves competitive with conventional reconstruction methods. Its modular hierarchical design extends naturally to additional dynamic factors and higher acceleration rates. The work targets reliable quantitative DCE-MRI analysis under high undersampling. Code is available at https://github.com/compai-lab/2026-GaborDCE-spieker, and the paper is on arXiv as 2608.18055.

Paper Overview

Field: Computer Vision (CV) / Medical Imaging
Authors: Veronika Spieker, Wenqi Huang, Cemre Ariyurek, Liam Timms, Daniel Rueckert, Onur Afacan, Julia A. Schnabel, Sila Kurugol
Published: 2026-08-18
arXiv: 2608.18055
Code: https://github.com/compai-lab/2026-GaborDCE-spieker

Abstract

Reliable quantitative analysis of dynamic contrast-enhanced MRI requires high-quality spatiotemporal reconstructions at high undersampling rates. Scan-specific reconstructions using Gaussian and Gabor primitives have shown promising results without the need for large training datasets, but have not addressed the additional dimension of dynamic contrast.

The authors propose a multi-dimensional, primitive-based framework for dynamic contrast-enhanced MRI reconstruction that disentangles the underlying anatomy, the dynamic contrast enhancement, and residual motion into separate temporal basis functions, thereby enabling a geometrical interpretation of the representation.

Key Findings

The architecture achieves performance competitive with conventional reconstruction methods, both in reconstruction quality and in the accuracy of extracted aortic and kidney enhancement curves.
The modular hierarchical design naturally extends to additional dynamic factors and higher acceleration rates.
The method requires no large training datasets, remaining fully scan-specific and unsupervised.

Summary in Chinese (from the forum post)

动态对比增强MRI的可靠定量分析需要高质量时空重建，特别是在高欠采样率下。该框架将解剖结构、动态对比增强和残余运动解耦到单独的时间基函数中，实现对表示的几何解释，并在重建质量和主动脉、肾脏增强曲线准确性上达到与传统方法相当的性能。

---

*Auto-collected on 2026-08-20*