Published January 31, 2025 | Version v2
Dataset Open

Quality-Assurance Package for the "Automated, Open-Source, Vendor-Independent Quality Assurance Protocol Based on the Pulseq Framework" Manuscript

Authors/Creators

  • 1. Division of Medical Physics, Department of Radiology, University Medical Center Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany
  • 2. School of Electrical Engineering, Chongqing University, Chongqing, China
  • 3. Core Facility MRDAC, Department of Radiology, University Medical Center Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany
  • 4. Department of Biomedical Engineering, University of Michigan, Ann Arbor, Michigan, USA

Description

Background

Neuroimaging research requires consistent image quality and temporal signal stability, especially for functional magnetic resonance imaging (MRI) studies that rely on detecting subtle blood-oxygen-level-dependent (BOLD) signal changes. Regular MR system performance monitoring is essential, especially for longitudinal and multi-site studies. This study aims to establish a robust quality assurance (QA) protocol to promote data comparability across scanner models, vendors, and sites, as well as over a prolonged period.

The manuscript titled "Automated, Open-Source, Vendor-Independent Quality Assurance Protocol Based on the Pulseq Framework" was submitted to the Special Issue Reproducibility and Quality Assurance of the Magnetic Resonance Materials in Physics, Biology and Medicine (MAGMA) journal.

This QA package proposed by the manuscript hosts materials for

  • all reconstructed images,
  • instruction for data acquisition,
  • instruction for image reconstruction,
  • instruction for post-processing,
  • example raw data and DICOM images, and
  • images and scripts for T1/T2 fitting.

The detailed information is listed below.

All reconstructed images

This directory contains all reconstructed images from the fBIRN phantom on three Siemens 3T scanners (Trio, Prisma.Fit, and Cima.X) and one GE (UHP) 3T scanner. It contains four sub-folders for each scanner. And each sub-folder contains (some of) the following sub-folders:

  • product_epi_ice: ICE-reconstructed product EPI images.
  • product_epi_gt: Gadgetron-reconstructed product EPI images.
  • pulseq_epi_ice: ICE-reconstructed Pulseq EPI images.
  • pulseq_epi_gt: Gadgetron-reconstructed Pulseq EPI images.
  • product_se_ice: ICE-reconstructed product spin-echo (SE) images.
  • product_se_gt: Gadgetron-reconstructed product SE images.
  • pulseq_se_ice: ICE-reconstructed Pulseq SE images.
  • pulseq_se_gt: Gadgetron-reconstructed Pulseq SE images.

Instruction for data acquisition

This directory includes the following documents:

  • write_QA_Tran_EPIrs.m to generate the QA_epi.seq file for EPI scans.
  • write_QA_Tran_T1.m: to generate the QA_T1.seq file for SE scans.
  • 20241122_QA_protocol_instruction_siemens.docx: standard operating procedure for QA measurements.
  • QA_record.xlsx: Excel sheet for the record of QA measurements.

Instruction for image reconstruction

Documents

  • pulseq2mrd_epi.m: convert GE Pulseq EPI raw data (.mat) to MRD raw data (.h5) using the LABEL information in the QA_epi.seq file.
  • pulseq2mrd_se.m: convert GE Pulseq SE raw data (.mat) to MRD raw data (.h5) using the LABEL information in the QA_T1.seq file.
  • siemens2mrd_epi.m: convert Siemens Pulseq EPI raw data (.dat) to MRD raw data (.h5) using the information in the .dat raw data.
  • default.xml: Gadgetron configuration file for SE image reconstruction. This document is already in the Gadgetron container: /opt/conda/envs/gadgetron/share/gadgetron/config/default.xml.
  • qc_epi.xml: Gadgetron configuration file for EPI image reconstruction, which is modified from the default epi.xml located in the Gadgetron container: /opt/conda/envs/gadgetron/share/gadgetron/config/.
  • specialCard_ICE.png: Special card setting for ICE online reconstruction.

Procedures for Gadgetron offline reconstruction

Step 1: Gadgetron installation (for more details, visit here)

  • Download and install Docker software. You may need to install/update the Windows Sub Linux (WSL) system for the Docker installation.
  • Open your terminal (Power shell with administrative privilege in Windows) and navigate to the folder you would like to map to the Gadgetron Docker container.
  • Run: docker run -t --name gt_latest --detach --volume ${pwd}:/opt/data ghcr.io/gadgetron/gadgetron/gadgetron_ubuntu_rt_nocuda:latest. If docker is not recognized, set docker to connect to C:\Program Files\Docker\Docker\resources\bin in the Environment Path in Windows. This will download and then launch the latest Gadgetron version in a Docker container. It will also mount your current folder as a data folder inside the container.
  • Run this command: docker exec -ti gt_latest /bin/bash. This will execute your Gadgetron container.

Step 2: Data preparation

  • Place your SE/EPI .dat/.h5 data in the mounted folder.
  • Run the command in Terminal: cd /opt/data to enter the mounted folder.

Step 3: MRD conversion

  • For Siemens data, you can convert the .dat data to MRD data by using Gadgetron. If Gsdgetron doesn't work (e.g. for XA EPI data), you can then use the Matlab script siemens2mrd_epi.m.
  • The command for Siemens SE data conversion: siemens_to_ismrmrd -f meas_MID*.dat -z 2 -o se_data.h5.
  • The command for Siemens EPI data conversion: siemens_to_ismrmrd -f meas_MID*.dat -z 2 -m IsmrmrdParameterMap_Siemens.xml -x IsmrmrdParameterMap_Siemens_EPI.xsl -o epi_data.h5.
  • For GE data, you can convert the .mat raw data to MRD data by using the Matlab scripts with the corresponding .seq files. For SE conversion: use pulseq2mrd_se.m with QA_T1.seq. For EPI conversion: use pulseq2mrd_epi.m with QA_epi.seq.

Step 4: Gadgetron reconstruction

  • SE reconstruction: gadgetron_ismrmrd_client -f se_data.h5 -c default.xml -o se_out.h5.
  • EPI reconstruction: first, put qc_epi.xml to the mounted folder and then copy it to the Gadgetron container: cp /opt/data/qc_epi.xml /opt/conda/envs/gadgetron/share/gadgetron/config/. Then, run the reconstruction: gadgetron_ismrmrd_client -f epi_data.h5 -c qc_epi.xml -o epi_out.h5.

Step 5: Load Gadgetron-reconstructed images (.h5)

  • Load SE .h5 images in Matlab:

filename = 'pulseq_se_out.h5' ;

info = hdf5info(filename) ;

address_data_1 = info.GroupHierarchy.Groups(1).Groups.Datasets(2).Name ;

pulseq_se_im = squeeze(double( hdf5read(filename, address_data_1) ) ) ;

pulseq_se_im = reshape(pulseq_se_im, [256, 256, 11, 2]) ;

  • Load EPI .h5 images in Matlab:

filename = 'pulseq_epi_out.h5';

info = hdf5info(filename) ;

address_data_1 = info.GroupHierarchy.Groups(1).Groups.Datasets(2).Name ;

pulseq_epi_im = squeeze(double( hdf5read(filename, address_data_1) ) ) ;

pulseq_epi_im = reshape(pulseq_epi_im, [64, 64, 27, 200]) ;

Procedures for ICE online reconstruction

Before executing the Pulseq-based sequences, you can enable ICE online Reconstruction following the procedures below:

  • Navigate to the Special Card (specialCard_ICE.png), set Data handling to ICE STD for NUMARIS/X (e.g. XA60A and XA61A), and ICE 2D for NUMARIS/4 (e.g. VB, VD, and VE).
  • Select Sum-of-Square for coil combination.
  • Be sure that the maximal pixel intensity does not violate the intensity threshold of 4096.

Instruction for post-processing

The example post-processing is based on the reconstructed images from Cima.X over five days.

Reconstructed images from Cima.X

Note: All se folders contain a structuralQuality_main.m to call the structuralQuality.m function for structural quality analysis. All epi folders contain a temporalQuality_main.m to call the temporalQuality.m function for temporal quality analysis.

  • product_epi_ice: ICE-reconstructed product EPI images.
  • product_epi_gt: Gadgetron-reconstructed product EPI images.
  • pulseq_epi_ice: ICE-reconstructed Pulseq EPI images.
  • pulseq_epi_gt: Gadgetron-reconstructed Pulseq EPI images.
  • product_se_ice: ICE-reconstructed product SE images.
  • product_se_gt: Gadgetron-reconstructed product SE images.
  • pulseq_se_ice: ICE-reconstructed Pulseq SE images.
  • pulseq_se_gt: Gadgetron-reconstructed Pulseq SE images.

QA analysis Matlab package: QA_functions

  • circfit.m: to find the center point and radius of the phantom.
  • makeCircleMask.m: to make a circular mask based on the center point and radius.
  • structuralQuality.m: to analyze the structural quality of the SE images.
  • temporalQuality.m: to analyze the temporal quality of the EPI images.

Post-processing procedures

  • Step 1: Add the QA_functions folder to your Matlab Path.
  • Step 2: Run the temporalQuality_main.m or structuralQuality_main.m script in each folder to produce the QA results of all reconstructed images inside the folder.
  • Step 3: Run the make_figure_epi.m and make_figure_se.m to produce some of the tables and figures used in the manuscript.

Example raw data and DICOM images

The data and DICOM images were acquired from Cima.X on the fBIRN phantom on 06.08.2024.

  • DICOM folder: contains the DICOM images for four EPI scans (the first two scans for warm-up) and two SE scans.
  • meas*.dat: Siemens raw data of two EPI scans for temporal quality analysis and two SE scans for structural quality analysis.
  • *data.h5 files: the ISMRMRD data of the four raw datasets.
  • *out.h5 files: the images reconstructed by Gadgetron.
  • *.nii: the NIFTI-format reconstructed images.
  • siemens2mrd_epi.m: to convert the Siemens EPI raw data to ISMRMRD data.
  • read_image.m: to convert the Gadgetron-reconstructed h5-format images to NIFTI-format images.

Images and scripts for T1/T2 fitting

This package includes DICOM images and T1/T2 fitting scripts for the fBIRN phantom. Images for T1 fitting were acquired using a product turbo spin echo sequence with an inversion recovery pulse (repetition time = 4000 ms, echo train length = 4). Images for T2 fitting were obtained using a product SE sequence (repetition time = 3500 ms). Both measurements were conducted on the Siemens Prisma.Fit 3T scanner on 05.06.2024.

  • T1 sub-folder: contains all DICOM images for T1 fitting with inversion recovery times of {50, 150, 300, 450, 600, 750, 900, 1050, 1200, 1350, 1500, 2200, 3000} ms.
  • T2 sub-folder: contains all DICOM images for T2 fitting with echo times of {7.5, 15, 30, 45, 60, 75, 90, 130, 200, 250} ms.
  • Do_T1fit.m: Matlab script for T1 fitting.
  • Do_T2fit.m: Matlab script for T2 fitting.

For more information regarding Pulseq and the workflow for data acquisition and image reconstruction, please visit our GitHub repositories: Pulseq Matlab software, Pulseq Tutorials, and Pulseq Rocks for the 2024 ISMRM Reproducibility Team Challenge.

If you need any further information or have any questions, please feel free to contact our Pulseq email address: pulseq.mr@uniklinik-freiburg.de.

Files

Files (25.4 GB)

Name Size Download all
md5:ca194f48088d3c0aba43be18fab20710
4.1 GB Download
md5:b983f890c3a238e37fb9e4086947a0cb
1.8 MB Download
md5:45133992456e1efe89035de235f570c8
20.4 GB Download
md5:7df3d89761da14ffb11b13e6f4f1cee4
495.6 kB Download
md5:5f39480e04ae2b307da7e074ac7df972
982.6 MB Download
md5:669daad831589f8e7b9285cde9720854
170.7 kB Download

Additional details

Funding

National Institutes of Health
R01 EB032378
National Institutes of Health
U24 NS120056

Dates

Available
2025-01-31

Software

Repository URL
https://github.com/HarmonizedMRI/qualityAssurance
Programming language
MATLAB , C++
Development Status
Active

References

  • Jezzard P, Matthews PM, Smith SM (2001) Functional Magnetic Resonance Imaging: An Introduction to Methods. Oxford University Press, Oxford, pp 4–34.
  • Friedman L, Glover GH (2006) Report on a multicenter fMRI quality assurance protocol. J Magn Reson Imaging 23:827–839.
  • Friedman L, Glover GH, The FBIRN Consortium (2006) Reducing interscanner variability of activation in a multicenter fMRI study: Controlling for signal-to-fluctuation-noise-ratio (SFNR) differences. Neuroimage 33:471–481.
  • Glover GH, Mueller BA, Turner JA, Van Erp TGM, Liu TT, Greve DN, Voyvodic JT, Rasmussen J, Brown GG, Keator DB, Calhoun VD, Lee HJ, Ford JM, Mathalon DH, Diaz M, O'Leary DS, Gadde S, Preda A, Lim KO, Wible CG, Stern HS, Belger A, McCarthy G, Ozyurt B, Potkin SG, FBIRN (2012) Function biomedical informatics research network recommendations for prospective multicenter functional MRI studies. J Magn Reson Imaging 36:39–54.
  • Jovicich J, Minati L, Marizzoni M, Marchitelli R, Sala-Llonch R, Bartrés-Faz D, Arnold J, Benninghoff J, Fiedler U, Roccatagliata L (2016) Longitudinal reproducibility of default-mode network connectivity in healthy elderly participants : a multicentric resting-state fMRI study. Neuroimage 124:442–454.
  • Yan C-G, Craddock RC, Zuo X-N, Zang Y-F, Milham MP (2013) Standardizing the intrinsic brain: towards robust measurement of inter-individual variation in 1000 functional connectomes. Neuroimage 80:246–262.
  • Kayvanrad A, Arnott SR, Churchill N, Hassel S, Chemparathy A, Dong F, Zamyadi M, Gee T, Bartha R, Black SE, Lawrence-Dewar JM, Scott CJM, Symons S, Davis AD, Hall GB, Harris J, Lobaugh NJ, MacQueen G, Woo C, Strother S, the ONDRI Founding Investigators, the CAN-BIND Investigators (2021) Resting state fMRI scanner instabilities revealed by longitudinal phantom scans in a multi-center study. Neuroimage 237:1–22.
  • Price RR, Axel L, Morgan T, Newman R, Perman W, Schneiders N, Selikson M, Wood M, Thomas SR (1990) Quality assurance methods and phantoms for magnetic resonance imaging: report of AAPM nuclear magnetic resonance Task Group No. 1. Med Phys 17:287–295.
  • Layton KJ, 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. Magn Reson Med 77:1544–1552.
  • Chen Q, Zijlstra F, Hucker P, Littin S, Zaitsev M (2024) Open-source, cross-platform workflow for MRI data acquisition and image reconstruction based on the Pulseq framework. Proc. 33rd Int. Soc. Magn. Reson. Med. Singapore, p 0948
  • Hansen MS, Sørensen TS (2013) Gadgetron: an open source framework for medical image reconstruction. Magn Reson Med 69:1768–1776.
  • Inati SJ, Naegele JD, Zwart NR, Roopchansingh V, Lizak MJ, Hansen DC, Liu C-Y, Atkinson D, Kellman P, Kozerke S, Xue H, Campbell-Washburn AE, Sørensen TS, Hansen MS (2017) ISMRM Raw data format: a proposed standard for MRI raw datasets. Magn Reson Med 77:411–421.
  • American College of Radiology (2018) Phantom test guidance for use of the large MRI phantom for the MRI accreditation program. 1–28.
  • American Association of Physicists in Medicine (AAPM) (2010) Acceptance testing and quality assurance procedures for magnetic resonance imaging facilities. Rep No 100 1–32.
  • Nielsen J-F, Noll DC (2018) TOPPE: a framework for rapid prototyping of MR pulse sequences. Magn Reson Med 79:3128–3134.
  • Chen Q https://github.com/HarmonizedMRI/qualityAssurance. In: Accessed August 22 2024.
  • National Electrical Manufacturers Association (2008) Determination of signal-to-noise ratio (SNR) in diagnostic magnetic resonance imaging. NEMA Stand. Publ. MS 1-2008. Rosslyn, pp 1–19
  • Gudbjartsson H, Patz S (1995) The rician distribution of noisy mri data. Magn Reson Med 34:910–914.
  • Delakis I, Petala K, De Wilde JP (2005) MRI receiver frequency response as a contributor to Nyquist ghosting in echo planar imaging. J Magn Reson Imaging 22:324–328.
  • Weisskoff RM (1996) Simple measurement of scanner stability for functional NMR imaging of activation in the brain. Magn Reson Med 36:643–645.
  • Simmons A, Moore E, Williams SCR (1999) Quality control for functional magnetic resonance imaging using automated data analysis and Shewhart Charting. Magn Reson Med 41:1274–1278.
  • Devor R, Chang T (1992) Statistical quality design and control: contemporary concepts and methods. Prentiss- Hall, New York
  • Shewhart WA (1931) Economic control of quality of manufactured product. D. Van Nostrand Co, New York