In vivo imaging of collagen fibre structures using dedicated low-field MRI
File(s)
Author(s)
Lanz, Harry
Type
Thesis or dissertation
Abstract
Anisotropic collagen-rich tissues such as the anterior cruciate ligament (ACL) and meniscus are notoriously difficult to assess with conventional magnetic resonance imaging (MRI). These tissues often appear hypointense, yet, at certain orientations within an MRI scanner they appear bright due to the "magic angle" effect.
The magic angle effect arises from the anisotropic collagen fibre microstructure of connective tissues and their orientation relative to the main magnetic field, B0. Signal intensity is minimised when fibres align parallel to B0, and is maximised when they are oriented at the magic angle, 54.7°, relative to B0. Although conventionally regarded as an artefact, this phenomenon can, if systematically exploited, provide quantitative, non-invasive information on collagen fibre microstructure.
To this end, a rotatable low-field MRI system was developed, enabling reorientation of B0 relative to a static subject. This facilitated the first in vivo implementation of Magic Angle Directional Imaging (MADI). MADI utilises multiple acquisitions at different orientations to induce magic angle effect intensity variations, which are then analysed to estimate collagen fibre orientations in tissues such as the ACL and meniscus.
This thesis presents the development of a complete MADI workflow tailored to this prototype system. A bespoke multi-stage registration pipeline was developed to align multi-orientation images and correct geometric distortions intrinsic to permanent magnet systems. An extensible semi-automatic segmentation framework was implemented to reduce operator burden whilst retaining segmentation precision. A significantly accelerated collagen fibre orientation estimation algorithm was developed, enabling optimisation of scan orientation selection using a genetic algorithm. Novel visualisation approaches were also developed to interpret the unique MADI data.
The present work culminated in the first successful in vivo application of MADI in healthy volunteers and patients with ACL and meniscal pathology, with results validated against arthroscopic findings, demonstrating the feasibility and potential for clinical translation of MADI.
The magic angle effect arises from the anisotropic collagen fibre microstructure of connective tissues and their orientation relative to the main magnetic field, B0. Signal intensity is minimised when fibres align parallel to B0, and is maximised when they are oriented at the magic angle, 54.7°, relative to B0. Although conventionally regarded as an artefact, this phenomenon can, if systematically exploited, provide quantitative, non-invasive information on collagen fibre microstructure.
To this end, a rotatable low-field MRI system was developed, enabling reorientation of B0 relative to a static subject. This facilitated the first in vivo implementation of Magic Angle Directional Imaging (MADI). MADI utilises multiple acquisitions at different orientations to induce magic angle effect intensity variations, which are then analysed to estimate collagen fibre orientations in tissues such as the ACL and meniscus.
This thesis presents the development of a complete MADI workflow tailored to this prototype system. A bespoke multi-stage registration pipeline was developed to align multi-orientation images and correct geometric distortions intrinsic to permanent magnet systems. An extensible semi-automatic segmentation framework was implemented to reduce operator burden whilst retaining segmentation precision. A significantly accelerated collagen fibre orientation estimation algorithm was developed, enabling optimisation of scan orientation selection using a genetic algorithm. Novel visualisation approaches were also developed to interpret the unique MADI data.
The present work culminated in the first successful in vivo application of MADI in healthy volunteers and patients with ACL and meniscal pathology, with results validated against arthroscopic findings, demonstrating the feasibility and potential for clinical translation of MADI.
Version
Open Access
Date Issued
2025-08-31
Date Awarded
2026-06-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Ristic, Mihailo
Sponsor
Wellcome Trust (London, England)
MedTechOne (Firm)
Grant Number
WT215908/Z/19/Z
PSR085_MEME
PSR079_MEME
Publisher Department
Department of Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)