Thesis

Developments in surface-enhanced spatially offset Raman spectroscopy for through-tissue imaging applications

Creator
Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2024
Thesis identifier
  • T16942
Person Identifier (Local)
  • 201862147
Qualification Level
Qualification Name
Department, School or Faculty
Abstract
  • In recent years, Raman-based techniques have been used extensively in bioanalytical research applications with the ultimate goal of creating platforms for medical diagnostics. Surface-enhanced spatially offset Raman spectroscopy (SESORS) is a powerful analytical technique that has emerged in an attempt to combine the signal enhancements offered by surface enhanced Raman scattering (SERS) with the subsurface probing in turbid media offered by spatially offset Raman spectroscopy (SORS). Using SESORS, it is possible to non-invasively retrieve subsurface spectra that originate from highly specific biofunctional SERS active nanotags inside diffusely scattering objects such as mammalian tissue. This has implications for a wide variety of applications where through barrier imaging and detection are used; the most obvious being in vivo clinical diagnostics. Fundamentally,however, SESORS imaging is not a well-understood technique. The diffusely scattering nature of most solid objects means that there is a tenuous link between the image or spectrum collected and the location of an inclusion, such as a clinically relevant disease state incubated with SERS active nanotags, at depth. This is true for the inclusion location in 3-dimensions. Despite that fact that the technique has already been applied to the numerous biological applications including drug delivery monitoring and in vivo cancer detection, detailed studies have not been completed that monitor the efficacy of SESORSimaging for the location of nanotags at depth. With this in mind, the focus of this research is to develop theoretical principals and experimental techniques that can be used to provide information about the location of SERS active nanotags within tissue, and to cultivate a better understanding of through barrier SORS and SESORS images for biomedical applications. Firstly, an internal calibration model is introduced for the prediction of the depth of SERS active nanotags embedded within a block of homogeneous porcine tissue using ratiometric analysis and a handheld SORS instrument. The modelutilises ratiometric analysis of the Raman intensities of the nanotags and the tissue barrier, or the relative contribution of the nanotags to through tissue spectra and is applicable down to depths where the nanotags can be discriminated using principal component analysis (PCA). It exploits the exponential decay in the relative contribution of the SERS active nanotags in offset spectra as the tissue barrier thickness is incrementally increased with a fixed spatial offset magnitude, and log-linear regression of this ratiometric quantity yields a linear calibration that correlates the Raman response of themulti-layered nanotag and tissue system with the buried depth of the nanotags. This approach is evaluated using a handheld backscattering SORS spectrometer in contact with a porcine tissue model and two different ‘flavours’ of SERS nanotags to demonstrate that it can be used equally on samples with different signatures, intensities, and depth penetration capabilities. Next, a fundamental question is addressed that is crucial to understanding SESORS imaging and implementing it in a clinical setting for in vivo diagnostic purposes, namely, can a SORS image be used to determine the exact locationof an object behind a barrier? To answer this, the effects of the spatial offset magnitude and geometry in the localisation of nanotags mixed with silica as an imaging target through tissue are investigated. Using an in-house built SORS system, experimental techniques are outlined that allow for the correct interpretation of SESORS images to ascertain the location of nanotag imaging templates in the x, y-imaging plane at depth. More specifically, the effect termed ‘linear offset induced image drag’ is presented, which refers to a spatial distortion in SESORS images caused by the magnitude and direction of an applied linear offset, and the need for an annular SORS collection geometry during imaging is highlighted to neutralise this asymmetric effect. Additionally,ratiometric SESORS imaging, an approach that utilises the relationshipbetween the magnitude of the spatial offset, the probed depth and ratiometric analysis of the nanotag and tissue Raman intensities, is introduced for the location of SERS active inclusions in 3-dimensions. Together, the separate approaches for localisation of nanotags in the x, y-imaging plane and the zaxis are combined to ultimately image and spatially discriminate between two optically distinct nanotag samples buried at different depths within the same 3-dimensional tissue model for the first time. Additionally, a technique for the identification of inclusion boundaries in 2-dimensional through tissue SERS and SESORS images is investigated. The result is a promising approach that allows for the sensitive delineation of the margins of SERS active NP inclusions non-invasively through tissue barriers. It utilises PCA of through tissue spectral datasets, that correspond to pixels within an image, to designate the pixels as active or inactive based on thepresence of SERS signatures indicative of NP inclusion location. We evaluate the ability of this method to delineate inclusion margins in through tissue SERS and SESORS images. This is done by validating images against an alternative binary image thresholding technique and rationalising the selected inclusion boundaries using a cumulative numerical integration technique that analyses image pixels based on multiple spectral regions. The technique, which we term “PCA-based inclusion margin delineation” is also tested by confirmation of previously established linear offset induced image drag effects caused by a point collection linear spatial offset, and finally, the influence of the SORS optical geometry and spatial offset magnitude on inclusion margins in multiple images collected on the same sample is investigated. Through the comparison of our proof-of-concept approach with traditional thresholding methods, and the application of the technique to observing previously established phenomena in SESORS images, this work aims to meet a previously unresolved issue in the field of through barrier Raman spectroscopy to bring itcloser to clinical use; namely, the non-invasive identification of SERS-active inclusion margins for comprehensive 2-dimensional localisation.Finally, depth determination of nanotags in tissue using the log-linearcalibration of the ratiometric quantity INP/ITis is performed in environments that add additional layers of complexity to the models developed earlier in the thesis. This includes investigating the effects of different tissue types, multiplexing, and the 2-dimensional position of the inclusion on the depth prediction model. New insights are gained into the parameters within the depth calibration model, as well as into the data processing and experimental techniques required to better understand and extract meaningful information from SESORS measurements.
Advisor / supervisor
  • Faulds, Karen
  • Graham, Duncan
Resource Type
DOI
Embargo Note
  • The digital version of this thesis is restricted to Strathclyde users only until 24th May 2029.

Las relaciones

Elementos