Quantitative Oblique Back-Illumination in the Study of Thick Biomedical Samples
Author(s)
Casteleiro Costa, Paloma Casteleiro
Advisor(s)
Editor(s)
Collections
Supplementary to:
Permanent Link
Abstract
The objective of this thesis is to explore the unique capabilities of quantitative oblique
back-illumination microscopy (qOBM), a novel tomographic, label-free, non-invasive, real-
time, and affordable quantitative phase imaging (QPI) technology. The aim is to develop
new qOBM-based optical and computational assays to warrant a more widespread use of
this technology for biomedical applications. qOBM overcomes QPI’s restriction to thin
samples, and enables high contrast and high-resolution quantitative phase imaging of thick
biomedical samples with cross-sectional and tomographic capabilities, providing valuable
morphological and biophysical information about the imaged specimen. In this work, we
first explore the application of qOBM in two clinical and biomedical areas, including the vi-
ability assessment of umbilical cord blood units for banking as well as surgery and pathol-
ogy assistance in the detection of brain tumor regions. Necessary modifications to the
optics and image analysis tools are presented in each of the aforementioned applications.
Secondly, we propose adapting qOBM for the non-invasive study of cellular and subcellu-
lar structural dynamics in three dimensional (3D) cell cultures. We again consider various
optical and computational modifications to the system required to capture fast biological
processes, and present data analysis pipelines to produce functional images of unlabeled
live samples. Lastly, we propose a deep learning single-capture approach to further sim-
plify and improve the system’s applicability.
Overall, the work presented in this dissertation seeks to establish the impact of qOBM
within the realm of biomedical optics. We do so by enhancing this technology’s accessi-
bility and effectiveness in a broad range of applications through modified optical designs
and advanced computational approaches. We expect this work to pave the way for the
development of novel label-free platforms for clinical and biomedical purposes.
Sponsor
Date
2023-06-15
Extent
Resource Type
Text
Resource Subtype
Dissertation