José Miguel Algarín1,2, Teresa Guallart-Naval1,2, José Borreguero3, Fernando Galve1,2, and Joseba Alonso1,2
1i3M, CSIC, Valencia, Spain, 2Universitat Politècnica de València, Valencia, Spain, 3Tesoro Imaging SL, Valencia, Spain
Synopsis
Keywords: Software Tools, Software Tools, Open-source, MaRCoS
Motivation: MaRCoS integrates hardware, firmware, and software for MRI scanner control, focusing on system reliability but overlooking user experience.
Goal(s): Develop MaRGA (MaRCoS Graphical Application), a user-friendly Graphical User Interface (GUI) for low-field MRI community, ensuring intuitive MaRCoS control and clinical environment compatibility.
Approach: Designed MaRGA with simplified panels and updated API, enabling features like DICOM image export, clinical protocol management, and streamlined image reconstructions.
Results: Tested MaRGA on 0.2 T and 72 mT scanners in lab and hospital settings, enhancing workflow efficiency and user satisfaction through customizable controls and improved overall experience.
Impact: The introduction of MaRGA, a user-friendly Graphical User Interface (GUI) for low-field MRI scanner control, enhances user experience and workflow efficiency with MaRCoS. MaRGA significantly improves the operation of low-field MRI scanners in both laboratory and clinical settings.
Introduction
The electronic control system or 'console' is pivotal to any MRI
scanner. A prominent open-source initiative is Magnetic Resonance Control
System (MaRCoS) [1, 2], a sophisticated and versatile open-source platform for
MRI control with high-performance attributes and low-cost. A relevant milestone
regarding MaRCoS is to develop a Graphical User Interface (GUI) that facilitates
scanner operation in laboratories and hospital environments. In this work, we
introduce MaRGA (MaRCoS Graphical Application) a new graphical user interface
(GUI) and application programming interface (API) [3] designed to ease the
interaction between users and MaRCoS.Methods
The MaRGA API and the GUI are both written in Python.
The former (Figure 1.a) includes methods to handle communication with MaRCoS, a
parent class that provides a structured framework for creating, executing, and
managing complex MRI sequences, and different child classes corresponding to
specific sequences. The GUI (Figure 1.b) has been developed using the PyQt5
library and integrates with the underlaying API, allowing users to communicate
with the MaRCoS server and to initiate sequences, adjust parameters, and
oversee data acquisition.
The GUI is based on three different windows. The
Session Window (Figure 1.c) serves as the initial point of interaction,
allowing users to input essential information related to the imaging session.
This window takes in details such as user credentials, patient information,
project identifiers, and other relevant metadata. In
the Main Window (Figure 1.d), users interact with the sequences and results. It
includes a menu bar and toolbars for interacting with the MaRCoS server,
running calibrations, visualizing or executing pulse sequences, saving or
loading input parameters, and creating or deleting protocols. Additionally,
there is a space to show previously executed or pending sequences and the
parameters used in those sequences. Finally, the Post-Processing Window can
be used to analyze and refine the acquired data, both in k-space and image
space.
To test the usability of MaRGA,
we conducted experiments in two distinct setups: a laboratory-based MRI scanner
used for preclinical dental imaging [4] and a hospital-based MRI scanner
specializing in in-vivo extremity imaging [5]. In the laboratory setup we run Pointwise Encoding Time Reduction with
Radial Acquisition (PETRA) [6], typically used for hard tissue imaging, in
a home-made phantom. In the system installed in the hospital we run a protocol including
an autocalibration and four RARE sequences with different contrasts in a
volunteer.Results
Figure 2 shows a screenshot of the
Main Window (top) and Post-Processing Window (bottom) with the results obtained
using PETRA in the laboratory-based scanner. By default, the PETRA
sequence is configured in the API to reconstruct the images with FFT after
regridding. Then, we use the postprocessing window to apply Algebraic
Reconstruction Technique (ART) [7], apply a BM4D [8] filter and upscale the
matrix size of the image.
In Figure 3, a screenshot of the Post-Processing
Window illustrates the four acquired images in the hospital-based scanner
within the protocol, both before (top) and after (bottom) post-processing. Here
we apply a BM4D filter to all images. For Inversion Recovery RARE and T1 RARE
sequences, we apply a cos-bell filter in k-space along the readout direction
before Fourier reconstruction, followed by BM4D.Discussion
MaRGA contains sequences for imaging (e.g. RARE, GRE or PETRA) as well
as sequences for calibration (Rabi flopping, shimming, Larmor frequency and
others). From MaRGA we can also run autocalibration sequences. To this end, we
set the field of view directly from a scout image, we select the region of
interest, we show the mean value and standard deviations to get a quick
estimation of the signal-to-noise ratio, and we develop protocols to generate
standardized sets of images from different sequences to make the GUI accessible
to non-expert users.
As we look forward, we plan to make MaRGA fully compatible with Pulseq
based on prior developments for MaRCoS [9]. Another area for future development
is to save raw data using the ISMRMRD standard [10] to ensure uniformity in
data representation and further enhances interoperability across various MRI
systems. An important aspect in the evolution of MaRGA lies in certifying the
software for clinical use to validate the reliability and accuracy of the
system and to provide confidence to users.Conclusion
In this work we present MaRGA, a graphical and application interface to
easily interact with MaRCoS, execute pulse sequences and visualize the results.Acknowledgements
Project funded by: the EU (EIC Transition, 101136407), EURAMET
(22HLT02), Spanish MICINN (PID2022-142719OB-C22), and the Valencian Innovation
Agency (INNVA1/2022/4, INNVA1/2023/30).
References
[1] T. Guallart‐Naval et al., “Benchmarking the performance of a low‐cost magnetic resonance control system at multiple sites in the open
MaRCoS community,” NMR Biomed, no. March, pp. 1–13, 2022, doi:
10.1002/nbm.4825.
[2] V. Negnevitsky et al., “MaRCoS,
an open-source electronic control system for low-field MRI,” Journal of
Magnetic Resonance, p. 107424, 2023, doi: 10.1016/j.jmr.2023.107424.
[3] https://github.com/yvives/PhysioMRI_GUI
[4] Algarín, J.
M., Díaz-Caballero, E., Borreguero, J., Galve, F., Grau-Ruiz, D., Rigla, J. P.,
Bosch, R., González, J. M., Pallás, E., Corberán, M., Gramage, C.,
Aja-Fernández, S., Ríos, A., Benlloch, J. M., & Alonso, J. (2020). Simultaneous imaging of
hard and soft biological tissues in a low-field dental MRI scanner. Scientific Reports, 10, 21470.
https://doi.org/https://doi.org/10.1038/s41598-020-78456-2
[5] Guallart-Naval,
T., Algarín, J. M., Pellicer-Guridi, R., Galve, F., Vives-Gilabert, Y., Bosch,
R., Pallás, E., González, J. M., Rigla, J. P., Martínez, P., Lloris, F. J.,
Borreguero, J., Marcos-Perucho, A., Negnevitsky, V., Martí-Bonmatí, L., Ríos,
A., Benlloch, J. M., & Alonso, J. (2022). Portable magnetic
resonance imaging of patients indoors, outdoors and at home. Scientific
Reports, 12. https://doi.org/10.1038/s41598-022-17472-w
[6] Grodzki, D. M., Jakob, P.
M., & Heismann, B. (2012). Ultrashort echo time imaging using pointwise
encoding time reduction with radial acquisition (PETRA). Magnetic Resonance
in Medicine, 67(2), 510–518. https://doi.org/10.1002/mrm.23017
[7] Gordon, R., Bender, R., & Herman,
G. T. (1970). Algebraic Reconstruction Techniques (ART) for three-dimensional
electron microscopy and X-ray photography. Journal of Theoretical Biology,
29(3), 471–481. https://doi.org/10.1016/0022-5193(70)90109-8
[8] Maggioni,
M., Katkovnik, V., Egiazarian, K. & Foi, A. Nonlocal transform-domain
filter for volumetric data denoising and reconstruction. IEEE Trans. Image
Process. 22, 119–133. https ://doi.org/10.5041/rmmj.10355 5 (2013).
[9] Layton, K. J., Kroboth,
S., Jia, F., Littin, S., Yu, H., Leupold, J., Nielsen, J. F., Stöcker, T.,
& Zaitsev, M. (2017). Pulseq: A rapid and hardware-independent pulse
sequence prototyping framework. Magnetic Resonance in Medicine, 77(4),
1544–1552. https://doi.org/10.1002/mrm.26235
[10] Inati, S. J., Naegele, J.
D., Zwart, N. R., Roopchansingh, V., Lizak, M. J., Hansen, D. C., Liu, C. Y.,
Atkinson, D., Kellman, P., Kozerke, S., Xue, H., Campbell-Washburn, A. E.,
Sørensen, T. S., & Hansen, M. S. (2017). ISMRM Raw data format: A proposed
standard for MRI raw datasets. Magnetic
Resonance in Medicine, 77(1),
411–421. https://doi.org/10.1002/mrm.26089