Efficient Volumetric and Continuous MRI

Summary

This presentation showcases advancements in efficient, continuous MR imaging techniques, focusing on quantitative imaging for brain analysis. The speaker highlights innovative approaches like magnetic resonance fingerprinting (MRF) and echo planar time-resolved imaging (EPTI), pushing the boundaries of spatial resolution, scan time, and reconstruction methods. These techniques enable high-quality multi-contrast imaging and multi-compartment modeling, allowing for rapid, reproducible scans with minimal motion artifacts. Applications extend to both low-field and high-field MR, with promising results in brain scans and multi-tissue compartment modeling. The goal is to enable faster, more precise clinical MR exams and multi-parameter mapping for brain research and diagnostics.

Takeaways

• 😀 The talk focuses on efficient, volumic, and continuous MR imaging techniques for rapid, quantitative brain exams.
• 😀 Advances in magnetic resonance fingerprinting (MRF) have significantly improved imaging efficiency, allowing for multi-contrast time series data acquisition with high signal-to-noise ratio (SNR).
• 😀 The use of volumic encoding and simultaneous multi-slice acquisition has allowed for better SNR and higher acceleration, reducing scan times.
• 😀 Machine learning and temporal low-rank subspace modeling have been used to improve reconstruction, reducing the number of images that need to be reconstructed for more efficient data processing.
• 😀 The lab has achieved whole-brain T1 and T2 maps at 1mm isotropic resolution in under 2 minutes at 3T, a substantial improvement over previous methods.
• 😀 The lab demonstrated strong repeatability across scanner vendors and time points, showing the potential for cross-site and long-term quantitative imaging.
• 😀 Fast reconstruction times and efficient contrast synthesis are key to enabling clinical deployment, as demonstrated with machine learning-based solutions.
• 😀 The use of echo planar time-resolved imaging (EPTI) allows for dynamic tracking of signal evolution, enabling detailed multi-parametric quantitative imaging.
• 😀 EPTI provides high repeatability with minimal bias across multiple scans, allowing for more reliable clinical imaging and motion robustness.
• 😀 The technology developed can be applied to both low-field (0.55T) and ultra-high-field (7T) MR, enabling efficient imaging with high spatial resolution and multi-compartment modeling of brain structures, such as myelin, axons, and extracellular water.

Q & A

What is the main focus of the talk?

The main focus of the talk is on efficient volumic and continuous magnetic resonance imaging (MRI) acquisition techniques, highlighting their application in quantitative imaging, multi-contrast MR exams, and multi-compartment modeling.

What technology has been developed to improve magnetic resonance fingerprinting (MRF)?

To improve MRF, the lab has focused on better volumic encoding, utilizing simultaneous multi-slice and 3D acquisitions, along with advances in reconstruction methods like temporal low-rank subspace and sequence optimization with camera outbound.

How has the efficiency of MRF improved over time?

The efficiency of MRF has improved significantly, with the ability to acquire whole brain data at one millimeter isotropic resolution in less than two minutes at 3T, which is around an order of magnitude faster than the original method, which took about 30 minutes.

How does the lab address the issue of repeatability across different MRI scanners and time points?

The lab has demonstrated repeatability across different MRI scanners and time points by performing scans on a healthy volunteer using Siemens and GE scanners, showing minimal differences in the results, even with a six-month gap between scans.

What role does machine learning play in improving MRF?

Machine learning is used to optimize the reconstruction pipeline, providing faster reconstructions and more realistic synthesized images, which improve the clinical applicability of MRF. For example, using a machine learning-based unit multi-branch approach, the lab has been able to generate more accurate clinical contrasts.

What challenges does low-field MRI present, and how can they be overcome?

Low-field MRI faces challenges like reduced B0 homogeneity and longer T2* times. However, these challenges can be leveraged by using longer spiral readouts and optimizing flip angle and TR to boost signal-to-noise ratio (SNR). Advanced reconstruction methods, such as subspace reconstruction, help overcome time-consuming issues in processing.

What is the role of echo planar time-resolved imaging (EPTI) in the work presented?

EPTI is used to track signal evolution over time, enabling more continuous and precise quantitative imaging. By capturing multiple temporal and spatial data points, EPTI allows for the creation of detailed multi-contrast and multi-compartment images with high resolution and repeatability.

What are some of the benefits of using EPTI in MRI scanning?

EPTI offers high-resolution imaging, reduced distortion and blurring, and the ability to track T1, T2*, T2 decay, and proton density changes in real-time. This technique is particularly advantageous for rapid scans, high repeatability, and motion robustness, making it suitable for clinical environments.

How does the lab aim to apply its techniques to low-field MRI scanners?

The lab aims to adapt its high-efficiency imaging techniques to low-field scanners (0.55T), targeting high isotropic resolution scans in a short time, such as whole brain scans in three to four minutes, by optimizing acquisition protocols and leveraging improved SNR and reduced B0 variation.

What future advancements does the lab hope to achieve with these MRI techniques?

The lab hopes to further push the boundaries of MRI by achieving even higher resolution multi-compartment quantitative mapping at ultra-low and ultra-high fields. This includes improving scan times, signal-to-noise ratios, and developing more efficient methods for clinical deployment.