KP-INR: A Dual-Branch Implicit Neural Representation Model for Cardiac Cine MRI Reconstruction

2025-08-16

AI Summary

Existing implicit neural representation (INR) methods for accelerated cardiac cine MRI reconstruction neglect local contextual features, limiting image quality. To address this, we propose KP-INR—the first k-space-oriented dual-branch INR model. One branch learns coordinate-based positional embeddings, while the other employs multi-scale convolutions to extract contextual features from target points and their neighborhoods; a cross-branch interaction mechanism enables effective feature fusion. KP-INR is trained end-to-end on unpaired undersampled data without requiring fully sampled ground-truth references. Evaluated on the CMRxRecon2024 dataset, KP-INR significantly outperforms state-of-the-art INR and conventional reconstruction methods, particularly in preserving myocardial motion details and structural consistency. These results demonstrate its strong potential for clinical deployment.

Technology Category

• Machine Learning: Kernel Methods
• Computer Vision: Multi-modal Vision
• Knowledge Representation and Reasoning: Knowledge Acquisition

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Cardiac Magnetic Resonance (CMR) imaging is a non-invasive method for assessing cardiac structure, function, and blood flow. Cine MRI extends this by capturing heart motion, providing detailed insights into cardiac mechanics. To reduce scan time and breath-hold discomfort, fast acquisition techniques have been utilized at the cost of lowering image quality. Recently, Implicit Neural Representation (INR) methods have shown promise in unsupervised reconstruction by learning coordinate-to-value mappings from undersampled data, enabling high-quality image recovery. However, current existing INR methods primarily focus on using coordinate-based positional embeddings to learn the mapping, while overlooking the feature representations of the target point and its neighboring context. In this work, we propose KP-INR, a dual-branch INR method operating in k-space for cardiac cine MRI reconstruction: one branch processes the positional embedding of k-space coordinates, while the other learns from local multi-scale k-space feature representations at those coordinates. By enabling cross-branch interaction and approximating the target k-space values from both branches, KP-INR can achieve strong performance on challenging Cartesian k-space data. Experiments on the CMRxRecon2024 dataset confirms its improved performance over baseline models and highlights its potential in this field.

Problem

• Improves cardiac cine MRI reconstruction quality
• Addresses limitations of current INR methods
• Enhances k-space feature representation learning

Innovation

• Dual-branch INR model for MRI reconstruction
• Combines positional embeddings and multi-scale features
• Cross-branch interaction enhances k-space approximation