Mazzolani, Andrea;
(2025)
Advanced Strain Retrieval for Optical Coherence
Elastography.
Doctoral thesis (Ph.D), UCL (University College London).
![[thumbnail of PhD_Thesis_Andrea_Mazzolani.pdf]](https://discovery.ucl.ac.uk/style/images/fileicons/text.png) |
Text
PhD_Thesis_Andrea_Mazzolani.pdf - Accepted Version Access restricted to UCL open access staff until 1 February 2026. Download (21MB) |
Abstract
Mechanical properties of tissue, such as elasticity and stiffness, are crucial health indicators. Techniques to measure these properties, known as elastography, include ultra-sound and magnetic resonance imaging. Recently, optical coherence tomography (OCT) has been utilized for elastography, resulting in optical coherence elastography (OCE), which offers higher sensitivity to tissue deformation and superior spatial resolution. Strain retrieval in OCE involves estimating tissue deformations to determine local strains. Conventional strain retrieval models use the OCT phase difference between unloaded and loaded signals to estimate strain. However, these models face significant limitations, mainly due to strong assumptions regarding the OCT phase difference. The motivation of my work was to develop accurate and robust OCE methods that do not rely on these assumptions and can overcome the limitations of conventional models. I first analyze the state-of-the-art OCT and OCE techniques, followed by an in-depth examination of the OCT phase difference used in conventional models. I demonstrated that the assumptions of conventional OCE models lead to inaccurate strain estimations, especially in tissues with highly heterogeneous stiffness, such as cancerous tissues. To overcome these limitations, I developed the mathematical formalism of the complex-speckle recorrelation model (CSR), which does not make these strong assumptions on the phase, and is highly robust. Subsequently, I used deep learning to create ElastoCNN, a more general OCE model that is faster and more robust than CSR, capable of retrieving strains under more general mechanical loadings. ElastoCNN was trained with simulated data, and I accelerated the OCT image simulation process by using analytical approximations and a novel technique I developed, denoted as multi-spectral regression (MSR). This enabled the generation of realistic OCE data and the successful training of ElastoCNN, which surpasses conventional models in robustness and accuracy, significantly enhancing the reliability of OCE.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Advanced Strain Retrieval for Optical Coherence Elastography |
| Language: | English |
| Additional information: | Copyright © The Author 2025. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Med Phys and Biomedical Eng |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10203988 |
Loading...
Loading...Loading...Loading...Archive Staff Only
 |
View Item |
