An open-source reference system for portable low-field MRI designed to support replication, reproducibility, benchmarking, and quantitative comparison is presented, aiming to support cross-site comparability, reproducible research, and collaborative development of future portable low-field MRI technologies.
Abstract
Despite its renewed attention, the pathway to point-of-care portable low-field MRI systems remains challenging, limiting adoption across research groups. Incomplete documentation limits reproducibility, causing redesign and complicating cross-system comparison. Moreover, non-standardized testing and characterization complicates ethical approval for clinical studies. We present an open-source reference system for portable low-field MRI designed to support replication, reproducibility, benchmarking, and quantitative comparison. The system is fully open source, based on a ~50 mT permanent magnet, and integrated with a cloud-native acquisition platform. Pulseq-based calibration, characterization, and imaging sequences assessed noise level, eddy currents, image-based SNR, and geometric accuracy. Quantitative T1, T2, and B0 mapping sequences were developed and evaluated against reference values. Initial results from independent replications at two sites were compared. The system reached a noise level of 1.4 relative to the thermal noise floor and short eddy-current decay constants of 27-32 us across all gradient channels. Geometric deviations were below 2 mm over the field of view. Image-based SNR were consistent between the independent replications. Measured T1 values closely matched specified values, with an average absolute error of 3.1(1.8)%, while T2 values were overestimated by 10.4(5.8)%. Simulations showed only marginal errors for both quantities, suggesting experimental error sources for T2 mapping. The reference system combines openly documented hardware, software, calibration procedures, phantoms, quantitative MRI, and simulation tools in a reproducible ecosystem, aiming to support cross-site comparability, reproducible research, and collaborative development of future portable low-field MRI technologies.
Low-field magnetic resonance imaging (MRI) provides an accessible, portable, and low-cost alternative to high-field scanners, expanding diagnostic imaging to point-of-care settings. However, widespread adoption is fundamentally hindered by a severely reduced signal-to-noise ratio (SNR). At low frequencies, radiofrequency (RF) coil conductor losses - rather than tissue sample losses - predominantly govern the system's total noise, making meticulous RF coil optimization critical to recovering image quality. This work presents an open-source, optimized solenoid head coil tailored for the 50 mT open-source scanner (OSII ONE v2.1). The paper validates production reproducibility across three independent international institutions and introduce an open-source connector with integrated digital circuitry for coil identification and DC or logic signals. Comprehensive benchtop measurements, Electromagnetic Interference (EMI) coupling analysis, Specific Absorption Rate (SAR) safety simulations, and phantom and human volunteer imaging confirm the design's efficacy, safety, and reproducibility. The results of the paper, when combined with the material provided in the open-source dedicated repositories, set the basis for a fully reliable and reproducible component for the open-source OSII ONE MRI scanner. In addition, the same optimization strategy and design material can be exploited for designing other RF coils for imaging of other body parts.
Umberto Zanovello, Julia Pfitzer, Ariane Ernst et al.· 1 citation
Brain volume change over time is an important imaging-based biomarker of disease. However, traditional MR scanners are associated with high direct and indirect costs both up front and over time, rendering them inaccessible to many around the world. Recently, ultra-low-field (64 mT) portable MR scanners have been introduced for clinical use and have been highly safe and informative for neurologic monitoring. Volumetrics present an important potential use for ultra-low-field MRI that has yet to be established. The aim of the present work was to examine qualitative similarities between volume measurements in adults and to address the questions of reproducibility through test-retest reliability and external validity using recent software updates to the Hyperfine Swoop system, versions 8.8.1 and 9.0.0. In twenty neurologically typical adults, structural volumes demonstrated high test-retest intraclass correlations regardless of software (>0.999-0.883). Regional intraclass correlations also were explored, with the lowest stability observed in structures that were relatively central and caudal, though more recent software provided relative improvements. Regardless of field strength or software, measurements were comparable to those established in growth charts. These findings constitute a crucial foundation for the clinical utility of 64 mT MRI in monitoring brain volume loss over time.
M. Stockbridge, Rex Wang, V. Neal et al.· Aperture Neuro· 0 citations
Magnetic resonance imaging (MRI) at low magnetic field strengths (under 1 T) has seen renewed interest, driven by technological advances that enhance accessibility, affordability, and versatility. Improvements in design, gradient performance, and artificial intelligence (AI)-assisted reconstruction enable diagnostic-quality neuroimaging at well under 1 T, disrupting the once-peerless dominance of high-field imaging. Modern low-field systems offer advantages in patient comfort, siting flexibility, and/or portability. Reduced infrastructure and safety constraints enable scanning in diverse settings, including outpatient, emergency, and intensive care environments. Lower susceptibility improves imaging near air-bone interfaces, metallic implants, and the skull base, while markedly lowering specific absorption rate (SAR) and acoustic noise. Clinical neurological applications extend across the brain, spine, and skull base and include pediatric or fetal imaging, where improved tolerance and reduced motion artifacts are particularly useful. Lower signal-to-noise ratio (SNR) and limited spatial resolution remain challenges. AI-enhanced reconstruction, optimized pulse sequences, and higher relaxivity of contrast agents help mitigate these drawbacks. Portable low-field MRI also enables point-of-care neuroimaging and intraoperative use, offering real-time insights to expedite diagnosis. However, it is critical to acknowledge that, while low-field applications have found important roles in terms of quality, speed, and performance, higher-field-strength scanners will remain the diagnostic standard for the foreseeable future.
B. N. Delman, D. Lefton, Mark Finkelstein et al.· Journal of computer assisted...· 0 citations
Magnetic resonance imaging (MRI) is a cornerstone of neurological care, serving as the gold standard for diagnosing pathologies, such as brain tumors and, together with computed tomography, stroke. However, the high capital costs, specialized infrastructure requirements such as expensive and bulky radiofrequency-shielding cages, and the operational complexity of conventional high-field scanners (1.5T and 3T) largely confine MRI to centralized imaging facilities. This scenario often leads to reliance on brain computed tomography or ultrasound in point-of-care settings, despite MRI's superior soft-tissue contrast for diagnosis and prognosis. In the past decade, there has been renewed interest in compact and simplified ultra-low-field (under 0.1T) MRI scanners, fueled by advances in engineering and computing. Here, we review the recent developments in ultra-low-field brain MRI, which enable imaging in open environments and demonstrate initial clinical applicability in point-of-care settings. We also envision future developments along 3 focus areas (ie, hardware, imaging protocols, and data-driven image formation and analysis) to address the current limitations of image quality and contrast in ultra-low-field brain MRI systems.
Ed X. Wu, Yujiao Zhao, Yilong Liu et al.· Stroke· 0 citations
Abstract Magnetic resonance imaging (MRI) is one of the most important imaging modalities in clinical diagnostics and biomedical research; however, its usability is significantly limited by the high technical and methodological variability across sites, scanners, and acquisition protocols. This lack of uniformity affects quantitative measurements, reduces their reproducibility, and complicates multicenter studies. In recent years, numerous initiatives and technical approaches have emerged, focusing on acquisition standardization, data harmonization, signal quality assessment, and validation of quantitative methods. This review summarizes current knowledge on neuroimaging data harmonization (e.g., ComBat), inter-scanner variability, radiomic feature repeatability, standardized QA procedures, and the challenges associated with integrating artificial intelligence into clinical workflows. Metrological frameworks such as the Quantitative Imaging Biomarker Alliance (QIBA) further emphasize the need for clearly defined measurands, reference methods, and reproducible acquisition conditions. Special attention is given to large dataset initiatives, preclinical standardization platforms, and open tools for MRI quality assessment. The review highlights the need for unified methodologies, transparent protocols, and robust validation frameworks that are essential for the reliable clinical translatability of MRI.
D. Gogola, P. Szomolányi· Measurement Science Review· 0 citations
Purpose: In advanced computational MRI techniques, acquisition and reconstruction techniques are jointly designed. For reproducibility, it is therefore important to provide an open implementation of both. At the same time, any use in a clinical environment usually requires a close integration with the MRI scanner. Ensuring long-time reproducibility and maintenance then poses additional challenges. In this work, we aim to provide a fully integrated open-source framework that can meet these demands. Methods: A software framework to develop pulse sequences is added to the BART toolbox. In addition, a vendor-specific driver sequence is developed that can be used to run the sequence on a clinical MRI scanner enabling online adjustment of all relevant sequence parameters. Using the Pulseq format, the exact same sequence can also be reproduced offline. As proof-of-concept, quantitative MRI methods for T1 and joint water/fat R2∗ mapping using radial FLASH and model-based reconstruction are implemented in the proposed framework. Consistency between online and offline acquisition is validated in phantom and in vivo experiments. Results: Quantitative MRI methods consisting of acquisition and reconstruction were successfully implemented in BART. Acquisition parameters and FOV can be adapted online on a clinical MRI system. Quantitative parameter maps from model-based reconstruction agree for online and offline regenerated Pulseq acquisitions. Conclusion: This work enables reproducibility of advanced computational MRI methods within a comprehensive end-to-end open-source framework.
Daniel Mackner, Philip Schaten, Markus Huemer et al.· arXiv.org· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduAug 17, 2026
A USAF cadet and a Lincoln Laboratory researcher found AI chatbots can help nontechnical service members produce viable software applications for their unique problems.
