Parallel transmission technologies for imaging the brain and body with 7 tesla MRI
<p>The push to higher-and-higher static magnetic field strengths for magnetic resonance imaging systems leads to the radiofrequency fields, that are used for image acquisition, becoming less-and-less uniform. This results in undesirable changes in image contrast which are not due to anatomy. I...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2024
|
| Subjects: |
| _version_ | 1824458777833766912 |
|---|---|
| author | Kent, JL |
| author2 | Hess, A |
| author_facet | Hess, A Kent, JL |
| author_sort | Kent, JL |
| collection | OXFORD |
| description | <p>The push to higher-and-higher static magnetic field strengths for magnetic resonance imaging systems leads to the radiofrequency fields, that are used for image acquisition, becoming less-and-less uniform. This results in undesirable changes in image contrast which are not due to anatomy. It is, however, possible to mitigate for this radiofrequency field non-uniformity using technology known as parallel transmission. Although, in order to provide a robust subject-specific improvement the non-uniform radiofrequency field must first be mapped in a process known as “B1+ mapping”.</p>
<p>To begin, I will go through some essential magnetic resonance physics and cover some of the key concepts used in this thesis. I will then introduce ultra-high field magnetic resonance imaging and discuss why it is of interest and some of the challenges it currently faces.</p>
<p>The first research chapter explores a new way of mapping this non-uniform radiofrequency field using a novel adaption to a previously used pulse sequence. This method, referred to as "Sandwich", was successful in producing absolute B1+ maps across the entire thorax at 7 T and in particular in the heart where typically many methods fail due to blood flow. Using extended phase graphs and Bloch simulations I investigated how this new method compares to those currently in use. I also investigated whether coil-cycling the acquisition of relative maps was beneficial and I found that coil-cycling is especially important when measuring 3D relative magnitude maps.</p>
<p>Mapping many individual transmit channels is time-consuming, hence a method to accelerate the acquisition of B1+ maps for the entire transmit array over a large volume was also explored. I did this by combining two methods known as "TxLR" and "B1TIAMO" with the Sandwich method mentioned previously. I showed that accurate B1+ maps can be reconstructed from low-resolution and sparse k-space data. I demonstrated this using pulse sequence capable of acquiring 3D single-channel maps for an 8-channel transmit array in only 14 seconds in the brain or 26 seconds in the body.</p>
<p>Finally, subject motion has the ability to degrade the performance of parallel transmission technology. Knowledge of the subject’s motion can help inform both the acquisition and reconstruction for motion correction. Motion detection can be achieved through various methods, but radiofrequency sensor-based techniques like parallel transmit scattering or pilot tone are of particular interest, primarily due to their low hardware requirements. Hence, the final research chapter explores these two competing radiofrequency sensor-based methods for motion detection. Both parallel transmit scattering and pilot tone showed good sensitivity to head motion and each method was able to accurately predict rigid body head motion in all six degrees of freedom using a linear regression model.</p> |
| first_indexed | 2025-02-19T04:31:17Z |
| format | Thesis |
| id | oxford-uuid:34d8b126-923b-428d-af30-b36796fbd958 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2025-02-19T04:31:17Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | oxford-uuid:34d8b126-923b-428d-af30-b36796fbd9582025-01-03T14:08:47ZParallel transmission technologies for imaging the brain and body with 7 tesla MRIThesishttp://purl.org/coar/resource_type/c_db06uuid:34d8b126-923b-428d-af30-b36796fbd958Ultra-high field MRIMRIMagnetic Resonance ImagingEnglishHyrax Deposit2024Kent, JLHess, AValkovič, LDragonu, I<p>The push to higher-and-higher static magnetic field strengths for magnetic resonance imaging systems leads to the radiofrequency fields, that are used for image acquisition, becoming less-and-less uniform. This results in undesirable changes in image contrast which are not due to anatomy. It is, however, possible to mitigate for this radiofrequency field non-uniformity using technology known as parallel transmission. Although, in order to provide a robust subject-specific improvement the non-uniform radiofrequency field must first be mapped in a process known as “B1+ mapping”.</p> <p>To begin, I will go through some essential magnetic resonance physics and cover some of the key concepts used in this thesis. I will then introduce ultra-high field magnetic resonance imaging and discuss why it is of interest and some of the challenges it currently faces.</p> <p>The first research chapter explores a new way of mapping this non-uniform radiofrequency field using a novel adaption to a previously used pulse sequence. This method, referred to as "Sandwich", was successful in producing absolute B1+ maps across the entire thorax at 7 T and in particular in the heart where typically many methods fail due to blood flow. Using extended phase graphs and Bloch simulations I investigated how this new method compares to those currently in use. I also investigated whether coil-cycling the acquisition of relative maps was beneficial and I found that coil-cycling is especially important when measuring 3D relative magnitude maps.</p> <p>Mapping many individual transmit channels is time-consuming, hence a method to accelerate the acquisition of B1+ maps for the entire transmit array over a large volume was also explored. I did this by combining two methods known as "TxLR" and "B1TIAMO" with the Sandwich method mentioned previously. I showed that accurate B1+ maps can be reconstructed from low-resolution and sparse k-space data. I demonstrated this using pulse sequence capable of acquiring 3D single-channel maps for an 8-channel transmit array in only 14 seconds in the brain or 26 seconds in the body.</p> <p>Finally, subject motion has the ability to degrade the performance of parallel transmission technology. Knowledge of the subject’s motion can help inform both the acquisition and reconstruction for motion correction. Motion detection can be achieved through various methods, but radiofrequency sensor-based techniques like parallel transmit scattering or pilot tone are of particular interest, primarily due to their low hardware requirements. Hence, the final research chapter explores these two competing radiofrequency sensor-based methods for motion detection. Both parallel transmit scattering and pilot tone showed good sensitivity to head motion and each method was able to accurately predict rigid body head motion in all six degrees of freedom using a linear regression model.</p> |
| spellingShingle | Ultra-high field MRI MRI Magnetic Resonance Imaging Kent, JL Parallel transmission technologies for imaging the brain and body with 7 tesla MRI |
| title | Parallel transmission technologies for imaging the brain and body with 7 tesla MRI |
| title_full | Parallel transmission technologies for imaging the brain and body with 7 tesla MRI |
| title_fullStr | Parallel transmission technologies for imaging the brain and body with 7 tesla MRI |
| title_full_unstemmed | Parallel transmission technologies for imaging the brain and body with 7 tesla MRI |
| title_short | Parallel transmission technologies for imaging the brain and body with 7 tesla MRI |
| title_sort | parallel transmission technologies for imaging the brain and body with 7 tesla mri |
| topic | Ultra-high field MRI MRI Magnetic Resonance Imaging |
| work_keys_str_mv | AT kentjl paralleltransmissiontechnologiesforimagingthebrainandbodywith7teslamri |