Tools and methods for low-field MR imaging and elastography

Yushchenko, Maksym. Tools and methods for low-field MR imaging and elastography. 2021, Doctoral Thesis, University of Basel, Faculty of Medicine.

[img] PDF
Restricted to Repository staff only until 1 June 2024.
Available under License CC BY-NC-SA (Attribution-NonCommercial-ShareAlike).


Official URL: https://edoc.unibas.ch/89874/

Downloads: Statistics Overview


Low-field magnetic resonance imaging (MRI) consists in performing MRI using relatively low-strength magnets (< 0.3 T) compared to the common clinical MR scanners (1.5 T - 3 T). Thanks to this feature, “low field” represents a different paradigm with respect to the conventional diagnostic workflow in human healthcare, which is currently constrained to specific radiology department environments in a hospital with fixed, large, yet mostly claustrophobic, expensive high-field MRI scanners requiring careful safety considerations when approaching them.
In fact, the lower magnetic field strength provides opportunities for small, compact, and mobile MR scanners, with reduced siting requirements and overall lower costs, which makes low-field MRI usage possible in settings that cannot be achieved by conventional scanners (such as a physician’s office, intensive-care units, or low-resource areas). Low-field MRI does not aim to compete with but rather to complement standard MR approaches and enable simpler, more efficient, or innovative use cases, such as organ-dedicated scanners, image-guided interventions, or imaging in conditions that would lead to severe susceptibility artifacts in stronger magnets (e.g. body parts containing ferromagnetic implants, organs with iron overload, lungs), as well as improved patient experience.
Recently, research and industrial groups have turned their attention to such long-neglected low-field opportunities after a long trend of MRI development with increased magnetic field strengths. Despite recent general technological advancements, low-field MRI is challenging because of the underlying MR physics: the magnetization of spins inside the body originating from the scanner’s main magnetic field is lower in a low-strength magnet, and the detected MR signal is consequently lower as well. If this signal is not sufficiently high with respect to the acquired noise, as represented by the signal-to-noise ratio (SNR), MRI scans may require diagnostically unfavorable compromises, such as longer duration or coarser image resolution. The main goal of low-field developments is thus to find all the possible means for improving SNR to keep scan time reasonably short for patient compliance and steadiness and to provide an image resolution suitable for the envisioned applications.
The present manuscript describes part of the PhD work at the Center for Adaptable MRI Technology focused on the development of efficient tools and methods for a non-commercial compact 0.1 T MRI scanner for human extremities. In particular, the described work includes the exploration of high-performance detectors for such a system, that would be capable of providing high SNR in the operating regime of a low magnetic field while remaining partially or entirely open (contrary to commonly used closed geometries). Together with the available scanner’s open access, such a feature can enable versatile imaging of variously flexed human extremities and the possibility of access with interventional tools for image-guided operations. This goal has been addressed first by optimizing a biplanar volume detector geometry with full three-side access, and then by integrating such a design into a quadrature detection configuration, where the combination of two decoupled acquisition channels provided an SNR gain of √2 factor. With such a high-performance detector and appropriately implemented MRI sequences, fine-resolution anatomical 3D images of a human ankle and elbow in flexed positions were successfully acquired in less than 10 minutes.
In parallel to the detector investigation, this PhD addressed the development of methods for MR elastography at low field. This advanced technique can estimate the mechanical properties of body tissues non-invasively, similar to manual palpation, to assess different pathological conditions that lead to changes of these properties, such as liver fibrosis or muscle atrophy. Low-field MR elastography can be potentially useful to perform such measurements when high-field scanner techniques fail or are not accurate due to artifacts or signal loss caused by iron accumulation in the organs or ferromagnetic implants proximity. However, no commercial elastography tools exist for low-field MRI systems, and their development in this setting has to tackle the severe constraints of elastography scans to avoid long acquisition times and poor SNR. On the one hand, this PhD investigated approaches for the evaluation of accuracy and reliability of MR elastography pipelines using numerical simulations and experimental acquisitions in a clinical scanner on a heterogeneous phantom made of silicone material with rheometric characterization. On the other hand, such phantoms were employed to develop custom tools and methods for low-field MR elastography. For this purpose, an innovative accelerated 3D acquisition strategy has been proposed to acquire the motion information of waves inside the body that are induced to perform MR elastography. Using the open quadrature detector, this approach allowed short scan times (<3 minutes) in human arms at 0.1 T, and thus, for the first time, it made in vivo MR elastography possible at magnetic fields lower than 1.5 T.
Advisors:Salameh, Najat
Committee Members:Sarracanie, Mathieu and Webb, Andrew
Faculties and Departments:03 Faculty of Medicine > Departement Biomedical Engineering > Imaging and Computational Modelling > Adaptable MRI Technology (Salameh)
UniBasel Contributors:Salameh, Najat
Item Type:Thesis
Thesis Subtype:Doctoral Thesis
Thesis no:14853
Thesis status:Complete
Number of Pages:iii, 175
Identification Number:
  • urn: urn:nbn:ch:bel-bau-diss148537
edoc DOI:
Last Modified:23 Nov 2022 05:30
Deposited On:22 Nov 2022 13:36

Repository Staff Only: item control page