Motion-robust free-running volumetric cardiovascular MRI

Overview

• Presents a new method called "compressive recovery with outlier rejection" (CORe) to improve the robustness of free-running volumetric cardiovascular MRI (CMR) to motion artifacts
• Compares CORe to traditional compressed sensing (CS), robust regression (RR), and an existing outlier rejection method using simulations and real-world CMR datasets
• Demonstrates that CORe can effectively suppress motion artifacts while maintaining or improving image quality and flow measurement reliability

Plain English Explanation

Cardiovascular MRI (CMR) is a powerful technique for imaging the heart and blood flow, but it can be sensitive to patient motion during the scanning process. This paper introduces a new method called "compressive recovery with outlier rejection" (CORe) that aims to make free-running volumetric CMR more robust to motion artifacts.

The key idea behind CORe is to model the motion-related outliers in the measured MRI data as a separate component that can be jointly estimated along with the underlying image. By enforcing the MR physics-guided "group sparsity" on this outlier component, CORe is able to effectively identify and suppress the motion artifacts while preserving the overall image quality.

Through extensive simulations and evaluations on real-world CMR datasets, the researchers demonstrate that CORe outperforms traditional compressed sensing (CS), robust regression (RR), and an existing outlier rejection method in terms of both image quality metrics and expert reader assessments. Particularly for 4D flow imaging, which captures the complex blood flow dynamics, CORe is shown to yield more reliable and consistent measurements, even in the presence of involuntary subject motion or exercise stress.

This work represents an important advancement in motion-robust CMR imaging, which has broad implications for improving the clinical utility and accessibility of this technology. By making CMR more resilient to patient motion, CORe can help enable wider adoption of free-running volumetric CMR techniques in both research and clinical settings.

Technical Explanation

The proposed CORe method models the motion-related outliers in the measured MRI data as an additive auxiliary variable. This auxiliary variable is then constrained using a "group sparsity" regularization term that enforces the MR physics-guided structure of the outliers.

The CORe algorithm jointly estimates the underlying image and the outlier component through an iterative optimization process. This approach allows CORe to effectively identify and suppress the motion artifacts while preserving the overall image quality.

The researchers evaluate CORe using two simulation studies and three real-world CMR datasets: 3D cine, 4D flow at rest, and 4D flow under stress. The simulation studies compare CORe to traditional CS, RR, and an existing outlier rejection method, demonstrating that CORe outperforms these baselines in terms of normalized mean square error and structural similarity index.

For the real-world 3D cine datasets, expert readers found that CORe was more effective in suppressing motion artifacts while maintaining or improving image sharpness compared to traditional CS. The 4D flow datasets further showed that CORe yields more reliable and consistent flow measurements, especially in the presence of involuntary subject motion or exercise stress.

Critical Analysis

The paper provides a comprehensive evaluation of the CORe method, using both simulations and real-world CMR datasets. The results demonstrate the effectiveness of CORe in suppressing motion artifacts and preserving image quality and flow measurement reliability.

However, the paper does not extensively discuss the limitations of the proposed method. For example, it is unclear how CORe would perform in the presence of more severe or complex motion patterns, or how the method's performance might vary across different patient populations or imaging protocols.

Additionally, while the authors provide implementation code and datasets on GitHub, the lack of a detailed description of the CORe algorithm in the paper may make it challenging for other researchers to fully understand and reproduce the method. A more thorough technical explanation, including the mathematical formulation and the iterative optimization procedure, would be beneficial for the broader research community.

Overall, the CORe method represents an important contribution to the field of motion-robust CMR imaging. Further research exploring the method's limitations and potential extensions could help solidify its place as a valuable tool for improving the clinical utility of free-running volumetric CMR techniques.

Conclusion

This paper presents a novel outlier mitigation method called "compressive recovery with outlier rejection" (CORe) that can significantly improve the robustness of free-running volumetric cardiovascular MRI (CMR) to motion artifacts. Through extensive simulations and evaluations on real-world CMR datasets, the researchers demonstrate that CORe outperforms traditional compressed sensing, robust regression, and an existing outlier rejection method in terms of both image quality and flow measurement reliability.

The CORe method's ability to effectively suppress motion-related artifacts while preserving overall image quality and flow data integrity has important implications for the wider adoption and clinical utility of free-running volumetric CMR techniques. By making CMR more resilient to patient motion, this work represents a significant advancement in the field and could help enable more accessible and reliable cardiovascular imaging for research and clinical applications.