MRI biomarkers of recovery for acute stroke
<p>Predicting stroke recovery is challenging. Imaging biomarkers might help address this unmet need. However, the majority of studies to date have been performed in the post-acute period or used simplistic biomarkers such as infarct volume and have largely ignored the dynamic processes that oc...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2021
|
| Subjects: |
| _version_ | 1826313007801565184 |
|---|---|
| author | Carone, D |
| author2 | Kennedy, J |
| author_facet | Kennedy, J Carone, D |
| author_sort | Carone, D |
| collection | OXFORD |
| description | <p>Predicting stroke recovery is challenging. Imaging biomarkers might help address this unmet need. However, the majority of studies to date have been performed in the post-acute period or used simplistic biomarkers such as infarct volume and have largely ignored the dynamic processes that occur immediately after stroke such as the contribution of oedema.</p>
<p>This Thesis set out to address these issues in a systematic manner in a cohort of patients recruited to a clinical imaging study at the John Radcliffe Hospital. Serial imaging data were acquired starting from the first hours following symptom onset.</p>
<p>The first part of this Thesis focused on improving measurement accuracy. A machine learning approach was developed and validated to segment stroke lesion and separately quantify infarct volume and anatomical distortion (oedema ± haemorrhagic transformation). An independent component analysis (ICA) based approach was used for the first time to denoise arterial spin labeling (ASL) perfusion data. Minimising the potential confounding factors present in the early stroke stages enabled the robust analysis and interpretation of the results obtained.</p>
<p>The second part of this Thesis used Topological Data Analysis (TDA) to analyse structural (DTI tractography) and functional data (rs-fMRI). TDA allows the study of higher order complex relationships between brain regions using spatially insensitive metrics. TDA metrics showed that both structural and functional connectivity were independent predictors of recovery. The changes in functional connectivity were independent of perfusion metrics, either BOLD or ASL derived. At 1-week, oedema was associated with symptom severity and recovery. Oedema was linked to presenting vasodilation and vasoconstriction, the latter providing supporting evidence for the glymphatic theory.</p>
<p>The results of this Thesis open the opportunity for further work to validate TDA metrics as biomarkers of recovery and to understand whether they may play a role in clinical trial design. In addition, a greater understanding of how post-stroke oedema reflects the balance of injury and repair mechanisms is needed to guide what treatment might be appropriate and when following acute stroke. These biomarkers could allow personalised treatment.</p> |
| first_indexed | 2024-03-07T07:06:46Z |
| format | Thesis |
| id | oxford-uuid:d0402ced-64b8-40d9-84b8-18bb198fbbb4 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-09-25T04:04:09Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | oxford-uuid:d0402ced-64b8-40d9-84b8-18bb198fbbb42024-05-14T08:51:09ZMRI biomarkers of recovery for acute strokeThesishttp://purl.org/coar/resource_type/c_db06uuid:d0402ced-64b8-40d9-84b8-18bb198fbbb4Cerebrovascular disease--PatientsEnglishHyrax Deposit2021Carone, DKennedy, J<p>Predicting stroke recovery is challenging. Imaging biomarkers might help address this unmet need. However, the majority of studies to date have been performed in the post-acute period or used simplistic biomarkers such as infarct volume and have largely ignored the dynamic processes that occur immediately after stroke such as the contribution of oedema.</p> <p>This Thesis set out to address these issues in a systematic manner in a cohort of patients recruited to a clinical imaging study at the John Radcliffe Hospital. Serial imaging data were acquired starting from the first hours following symptom onset.</p> <p>The first part of this Thesis focused on improving measurement accuracy. A machine learning approach was developed and validated to segment stroke lesion and separately quantify infarct volume and anatomical distortion (oedema ± haemorrhagic transformation). An independent component analysis (ICA) based approach was used for the first time to denoise arterial spin labeling (ASL) perfusion data. Minimising the potential confounding factors present in the early stroke stages enabled the robust analysis and interpretation of the results obtained.</p> <p>The second part of this Thesis used Topological Data Analysis (TDA) to analyse structural (DTI tractography) and functional data (rs-fMRI). TDA allows the study of higher order complex relationships between brain regions using spatially insensitive metrics. TDA metrics showed that both structural and functional connectivity were independent predictors of recovery. The changes in functional connectivity were independent of perfusion metrics, either BOLD or ASL derived. At 1-week, oedema was associated with symptom severity and recovery. Oedema was linked to presenting vasodilation and vasoconstriction, the latter providing supporting evidence for the glymphatic theory.</p> <p>The results of this Thesis open the opportunity for further work to validate TDA metrics as biomarkers of recovery and to understand whether they may play a role in clinical trial design. In addition, a greater understanding of how post-stroke oedema reflects the balance of injury and repair mechanisms is needed to guide what treatment might be appropriate and when following acute stroke. These biomarkers could allow personalised treatment.</p> |
| spellingShingle | Cerebrovascular disease--Patients Carone, D MRI biomarkers of recovery for acute stroke |
| title | MRI biomarkers of recovery for acute stroke |
| title_full | MRI biomarkers of recovery for acute stroke |
| title_fullStr | MRI biomarkers of recovery for acute stroke |
| title_full_unstemmed | MRI biomarkers of recovery for acute stroke |
| title_short | MRI biomarkers of recovery for acute stroke |
| title_sort | mri biomarkers of recovery for acute stroke |
| topic | Cerebrovascular disease--Patients |
| work_keys_str_mv | AT caroned mribiomarkersofrecoveryforacutestroke |