Additional sampling directions improve detection range of wireless radiofrequency probes

Malte Hoffmann, Marius Mada, T. Adrian Carpenter, Stephen J. Sawiak, Guy WilliamsMagnetic 2015 · DOI

TL;DR

This paper introduces a novel method for improving the accuracy and range of wireless radiofrequency (RF) markers used in MRI to track patient head motion. By increasing the number of sampling directions for these markers, the researchers significantly enhance the system's ability to correct for larger and more complex head movements, a critical challenge in obtaining high-quality MRI data.

What Problem Does This Paper Address?

Magnetic Resonance Imaging (MRI) is a powerful tool for visualizing the human brain, but its utility is often hampered by patient motion. Even slight movements during an MRI scan can introduce artifacts, degrading image quality and potentially leading to misinterpretations. While prospective motion correction techniques aim to address this by adjusting the scan in real-time, existing methods using wireless RF markers have limitations. Specifically, these methods rely on one-dimensional (1D) "navigator" readouts to determine marker positions. When the head rotates significantly, the signals from these markers can become ambiguous or overlap, making it impossible to accurately determine their location and thus correct for motion. This paper tackles the problem of extending the motion correction range of these wireless RF markers, particularly for patients who are prone to larger movements.

Methods

The core innovation presented in this paper is the "directions" method, which expands upon existing wireless RF marker tracking techniques. Instead of relying on a minimal set of 1D navigators (typically along the x, y, and z axes), the researchers incorporate additional sampling directions. These extra directions are strategically chosen to provide more comprehensive spatial information about the markers.

The authors investigated including up to 21 sampling directions, distributed across a sphere. This increased dimensionality helps to resolve ambiguities that arise when marker projections overlap in the simpler 1D case. When the head rotates, the projected signals from the markers can become difficult to distinguish. By acquiring more projections from different angles, the system can better triangulate the true 3D position of each marker, even under significant rotation.

The paper also explores the trade-off between readout resolution and speed. They found that reducing the resolution of the navigator readouts could significantly speed up the process without substantially compromising accuracy, offering a practical way to mitigate the increased acquisition time from more sampling directions.

Datasets

The study utilized data from a volunteer participant undergoing MRI scans on a 3T Siemens scanner. The researchers acquired data using various pulse sequences, including 3D GRE and 3D FSE, to test the performance of their method under different imaging conditions and to evaluate RF safety. No new benchmark datasets were introduced; rather, the evaluation was performed by comparing the marker-based motion estimates against established retrospective motion correction techniques like SPM8.

Key Insights and Ideas

  • More Directions = More Robustness: The fundamental insight is that increasing the number of sampling directions for RF markers dramatically improves their ability to disambiguate marker positions, especially during rotations. This directly translates to a wider range of motion that can be reliably tracked.
  • The "Correspondence Problem" is Key: The paper highlights the "correspondence problem" – the challenge of matching detected peaks in the navigator data to the correct physical markers. Additional sampling directions are crucial for resolving this ambiguity.
  • Resolution vs. Speed Trade-off is Manageable: The authors demonstrate that reducing navigator readout resolution is a viable strategy to regain speed lost by acquiring more directions, making the method more practical for real-time applications.
  • Real-World Motion is Significant: The study emphasizes the need for robust motion correction by showcasing patient motion exceeding 17 degrees in the z-direction, a level that challenges existing methods.

Key Results

  • Extended Rotation Range: The "directions" method demonstrated successful tracking of head rotations within ranges of -19.2° to 34.4° (x-axis), -2.7° to 10.0° (y-axis), and -60.9° to 70.9° (z-axis).
  • Improved Accuracy: In the presence of excessive head motion, the deviation of marker estimates from SPM8 was reduced by 17.1% compared to existing three-projection methods.
  • Time Efficiency: While adding seven directions increased readout time by a factor of 3.3, reducing resolution allowed for significant time savings without compromising accuracy.
  • RF Safety: The wireless probes showed minimal temperature increases (<0.5°C) and caused a maximum change of 1.3% in B1 mapping, indicating good RF safety.

Limitations and Future Work

  • Signal Fade: A limitation is that marker signals can fade if their axis becomes aligned with the main magnetic field (B0), leading to tracking failure.
  • Dislodging Markers: While not experienced in their volunteer testing, the possibility of large motions dislodging the markers remains a potential issue.
  • Comparison with Other Methods: The authors suggest that a detailed comparison with other advanced motion correction techniques (e.g., vNavs, fatNavs) is an important next step.
  • Real-time Implementation: While the processing time is reduced, further optimization for real-time prospective correction, potentially through compiled routines, is mentioned.
  • Exploiting Redundancy: Future work could focus on better exploiting the redundancy in the data from multiple directions to further improve tracking accuracy.

Summary updated

Abstract

Including additional sampling directions extends the range in which markers can be used, for patients who move a lot. Magn Reson Med 76:913-918, 2016. © 2015 The Authors. Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
BibTeX
@inproceedings{hoffmann2015additional,
  author = {Malte Hoffmann and Marius Mada and T. Adrian Carpenter and Stephen J. Sawiak and Guy Williams},
  title = {{Additional sampling directions improve detection range of wireless radiofrequency probes}},
  booktitle = {Magnetic},
  year = {2015},
  doi = {10.1002/mrm.25993},
}
Figure