Automated, open-source, vendor-independent quality assurance protocol based on the Pulseq framework

Area of Science:

• Neuroimaging
• Medical Physics
• Data Science

Background:

• Consistent image quality and signal stability are critical for neuroimaging, especially functional MRI (fMRI) which detects subtle BOLD signal changes.
• Regular MR system performance monitoring is vital for longitudinal and multi-site studies to ensure data reliability.

Purpose of the Study:

• To establish a robust, open-source, and vendor-independent quality assurance (QA) protocol.
• To enhance data comparability across different days, scanner versions, vendors, and research sites.

Main Methods:

• Implemented an open-source, vendor-independent QA protocol using Pulseq for standardized data acquisition.
• Utilized ISMRMRD/Gadgetron for harmonized image reconstruction and an automated post-processing pipeline for quality evaluation.
• Tested the protocol on multiple Siemens and GE 3T scanners across different sites and software versions, including phantom studies.

Main Results:

• The vendor-independent QA protocol yielded image quality comparable to vendor-based protocols.
• Demonstrated similar day-to-day repeatability to vendor-based protocols on Siemens scanners.
• Achieved high inter-day repeatability on GE scanners.

Conclusions:

• Successfully developed and implemented an open-source, vendor-independent QA protocol with an automated post-processing pipeline.
• The protocol proved feasible and repeatable across various conditions: different days, system versions, vendors, and sites.
• This QA approach is valuable for improving standardization in multi-site and longitudinal neuroimaging studies.