Abstract
BackgroundMedical images, e.g. of bilateral watershed (WS) in hypoxic ischaemic injury
(HII), are difficult to communicate - novel demonstration mechanisms are
needed.
Aims
To investigate patterns of injury in term HII on delayed MRI and develop and
validate novel methods of demonstration.
Objectives
- Demonstrate paucity of literature on delayed MRI in HII and
understanding of WS injury and introduce novel demonstration
methods.
- Quantify patterns of injury through baseline MRI interpretation.
- Describe method of flat earth map creation.
- Validate maps by testing laypeople’s ability to identify pathology.
- Improve segmentation and preparation of MRIs for 3D printing.
- Print 3D models and check fidelity.
- Validate 3D prints by testing laypeople’s ability to identify pathology.
Methods
Systematic literature review and focused review of the WS pattern of HII was
performed. Baseline MRI reads were performed to determine frequencies
and patterns of injury in a selected cohort. Two techniques for providing a
global perspective of the cortex were developed – Mercator flat-earth maps
and 3D prints. These were tested on untrained non-radiologists and lay
people.
Results
Paucity of literature demonstrated a knowledge gap. Most frequent patterns
of injury were basal-ganglia-thalami/ watershed (BGT/WS) (29.8%) and BGT
alone (24.2%). High statistical correlations were shown for the following
injuries:
- Thalami, basal ganglia, cortex.
- Corticospinal tracts, basal ganglia, thalami and hippocampi.
- Watershed, ulegyria, splenium of the corpus callosum, thalami.
Accurate flat earth maps and 3D models were created and validated in
demonstrating pathology to radiologists and laypeople – flat earth maps
(sensitivity 67%; specificity 80%); 3D prints (sensitivity 56.1%; specificity
55.3%).
Conclusion
Knowledge gaps regarding WS HII on delayed MRI were identified. Statistical
relationships between injured structures re-enforce known patterns of HII.
Flat earth maps and 3D prints were validated in demonstrating watershed
injury accurately to lay people.
| Date of Award | 13 May 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Savvas Andronikou (Supervisor), Jade Thai-Goodson (Supervisor) & Elavazhagan Chakkarapani (Supervisor) |
Cite this
- Standard