Quantitative methods for MRI-microscopy comparisons
<p>Magnetic resonance imaging (MRI) is a powerful tool for the in-vivo diagnosis and assessment of neurodegenerative disorders. While numerous MRI techniques have yielded parameters sensitive to microstructural changes in the brain, MRI parameters are infamously non-specific. Different changes...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2023
|
| Subjects: |
| _version_ | 1797113270518153216 |
|---|---|
| author | Kor, DZL |
| author2 | Miller, K |
| author_facet | Miller, K Kor, DZL |
| author_sort | Kor, DZL |
| collection | OXFORD |
| description | <p>Magnetic resonance imaging (MRI) is a powerful tool for the in-vivo diagnosis and assessment of neurodegenerative disorders. While numerous MRI techniques have yielded parameters sensitive to microstructural changes in the brain, MRI parameters are infamously non-specific. Different changes in brain tissue structure may result in the same change in MRI signal, making it challenging to pinpoint the exact source of these signal changes.</p>
<p>Biophysical modelling aims to achieve better biological specificity by relating dMRI signals to biologically interpretable tissue parameters. Yet, the fitting of these complex models with many parameters to unremarkable dMRI signals is challenging, often resulting in a degeneracy where multiple combinations of parameters explain the dMRI signal equally well.</p>
<p>Conversely, microscopy offers high biological specificity by targetting specific aspects of microstructure. By acquiring and relating MRI and microscopy metrics for the same tissue section, one can therefore leverage microscopy’s specificity to elucidate the microstructural basis of MRI signal change. However, microscopy is incredibly time-intensive, restricting examination to a few tissue sections at a time. A significant gap remains in the lack of a standardised pipeline for MRI-microscopy comparisons, with existing methods necessitating substantial manual intervention.</p>
<p>This thesis delves into methods that enhance MRI-microscopy comparisons. Specifically, we introduce an automated pipeline that rapidly and reliably extracts multiple quantitative microscopy parameters from sections that are histologically-stained. Utilising this pipeline alongside high-quality MRI-microscopy co-registrations, we performed whole-slide voxelwise comparisons between multimodal MRI- and microscopy-derived metrics. Finally, we present an alternative analysis method designed to relate degenerate biophysical model parameters to a continuous metric (e.g. from microscopy).</p>
<p>Overall, the techniques outlined in this thesis are intended to facilitate a more precise interpretation of microstructural change from MRI parameters and encourage a more systematic approach to processing microscopy data when relating them to MRI data.</p> |
| first_indexed | 2024-04-23T08:26:09Z |
| format | Thesis |
| id | oxford-uuid:a93d7034-ba8f-4e42-ae5a-712c66ece4c6 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-04-23T08:26:09Z |
| publishDate | 2023 |
| record_format | dspace |
| spelling | oxford-uuid:a93d7034-ba8f-4e42-ae5a-712c66ece4c62024-04-18T15:19:57ZQuantitative methods for MRI-microscopy comparisonsThesishttp://purl.org/coar/resource_type/c_db06uuid:a93d7034-ba8f-4e42-ae5a-712c66ece4c6Stains and staining (Microscopy)MicroscopyDiffusion magnetic resonance imagingEnglishHyrax Deposit2023Kor, DZLMiller, KHoward, AJbabdi, S<p>Magnetic resonance imaging (MRI) is a powerful tool for the in-vivo diagnosis and assessment of neurodegenerative disorders. While numerous MRI techniques have yielded parameters sensitive to microstructural changes in the brain, MRI parameters are infamously non-specific. Different changes in brain tissue structure may result in the same change in MRI signal, making it challenging to pinpoint the exact source of these signal changes.</p> <p>Biophysical modelling aims to achieve better biological specificity by relating dMRI signals to biologically interpretable tissue parameters. Yet, the fitting of these complex models with many parameters to unremarkable dMRI signals is challenging, often resulting in a degeneracy where multiple combinations of parameters explain the dMRI signal equally well.</p> <p>Conversely, microscopy offers high biological specificity by targetting specific aspects of microstructure. By acquiring and relating MRI and microscopy metrics for the same tissue section, one can therefore leverage microscopy’s specificity to elucidate the microstructural basis of MRI signal change. However, microscopy is incredibly time-intensive, restricting examination to a few tissue sections at a time. A significant gap remains in the lack of a standardised pipeline for MRI-microscopy comparisons, with existing methods necessitating substantial manual intervention.</p> <p>This thesis delves into methods that enhance MRI-microscopy comparisons. Specifically, we introduce an automated pipeline that rapidly and reliably extracts multiple quantitative microscopy parameters from sections that are histologically-stained. Utilising this pipeline alongside high-quality MRI-microscopy co-registrations, we performed whole-slide voxelwise comparisons between multimodal MRI- and microscopy-derived metrics. Finally, we present an alternative analysis method designed to relate degenerate biophysical model parameters to a continuous metric (e.g. from microscopy).</p> <p>Overall, the techniques outlined in this thesis are intended to facilitate a more precise interpretation of microstructural change from MRI parameters and encourage a more systematic approach to processing microscopy data when relating them to MRI data.</p> |
| spellingShingle | Stains and staining (Microscopy) Microscopy Diffusion magnetic resonance imaging Kor, DZL Quantitative methods for MRI-microscopy comparisons |
| title | Quantitative methods for MRI-microscopy comparisons |
| title_full | Quantitative methods for MRI-microscopy comparisons |
| title_fullStr | Quantitative methods for MRI-microscopy comparisons |
| title_full_unstemmed | Quantitative methods for MRI-microscopy comparisons |
| title_short | Quantitative methods for MRI-microscopy comparisons |
| title_sort | quantitative methods for mri microscopy comparisons |
| topic | Stains and staining (Microscopy) Microscopy Diffusion magnetic resonance imaging |
| work_keys_str_mv | AT kordzl quantitativemethodsformrimicroscopycomparisons |