UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On the functional classification of chemical and biological systems by the acquisition and multivariate analysis of optical, spectroscopic and electronic images Melo, Luke


Leaps forward in technology now yield volumes of data so immense they surpass the capacity of humans to interpret. At the core of data science, automated algorithms reliably transform enormous sums of raw data into meaningful information. A diversity of data in the form of vibrational spectra, microscope images and electrocardiograms (ECGs) appear very different at face value, however the algorithms used to process these data structures share many common threads. This first part of this thesis presents the progress made on the development of a novel microscope integrating collinear Raman and interferometric scattering (iSCAT) microscopies. iSCAT directly images sub-diffraction limited nanoscopic particles without the need for extraneous labelling. This overcomes one of the main drawbacks of fluorescence microscopy, however iSCAT fails to provide any chemical specificity. The integration of Raman microscopy overcomes this hurdle by probing the vibrational energy structure of the target analyte. The second part of this thesis presents the progress made on the development of a novel, non-invasive, and inexpensive diagnostic methodology for risk of sudden cardiac death (SCD). Despite the prevalence of heart disease as the leading cause of death worldwide, over half of SCD incidents remain unexplained. Brugada Syndrome (BrS) is an inherited arrhythmic cardiomyopathy which often lies dormant until manifests for the first time as SCD. In some parts of the world a physician can administer a sodium blocker challenge to reveal the signature of BrS in an ECG, however its use may cause life-threatening proarrhythmic effects. In collaboration with the IRCCS Policlinico San Donato, we have introduced a novel non-invasive methodology to diagnose BrS from the analysis of an ECG using a deep neural network.

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International