Lipidomics is the comprehensive study of cellular lipids on a large scale based on analytical chemistry principles and technological tools. It provides in-depth understanding of subtle alterations of lipids in cells in response to internal and/or external stimuli, such as environmental stress, diseased condition, drug treatment, and genetic mutation. Mass spectrometry (MS) has been recognized as a vital technique for lipidomics, and MALDI-MS is particularly attractive due to the speed, sensitivity, resolution and throughput benefits. Single-cell lipidomics addresses lipidome characterization at a single cell level, but its study has been largely hindered by complex structural diversity of lipids and limited sample amounts. This thesis describes the methodology development specifically for lipidomic study at the single cell level with a novel, high performing MALDI-MS platform enhanced by plasmonic substrates. The lipid profiling and analysis were applied to a series of cells, including algae, bacteria, and virus from clinical samples.Chapter Two describes a microchip-based MALDI-MS method to investigate the cytotoxic effects of herbicides on algae through single-cell lipid profiling in combination with machine learning (ML). The exposure of algal species Selenastrum capricornutum to different common herbicides and the resulting cytotoxic behaviors under stressed conditions were characterized. A lipid library for S. capricornutum has been established. Machine learning algorithms were applied to the classification of herbicide impact and identification of lipid species affected by the chemical exposure, leading to accurate identification of previously hidden cytotoxic differences. Chapter Three describes the use of gold nanofilm/MALDI-MS for untargeted lipidomic analysis of E. coli. Lipid profiling was performed with intact cells to understand the cellular response to antibiotic treatment (colistin), and statistical models were utilized for classification, variability, and reproducibility assessment. A number of lipids were evaluated as potential biomarkers for indicating gram-negative bacterial response to colistin. Chapter Four describes a MALDI-MS study with photosensitizer-induced signal enhancement for bacterial lipidomic profiling at both bulk and single cell level. Unique metabolite profiles were constructed by MS measurements and assisted by various statistical models including PLS-DA and LDA. The combination of photosensitizer and fluorescence localization allows discrimination of different bacterial species, providing a potential approach for bacterial identification without a proliferation process. Chapter Five describes the application of a novel plasmonic chip (Al-chip) with MALDI-MS for virus analysis through lipid profiling. Al-chip delivered higher sensitivity and thus better MS signals than gold. Clinical samples of COVID-19 nasal swabs were investigated by running lipid metabolite profiling to evaluate the feasibility of high throughput screening of viral infection. A comprehensive data processing and statistical analysis was utilized to reveal key changes among different lipid metabolites and evaluate their potential as diagnostic markers identifying positive and negative samples for rapid COVID-19 screening.

Due to the higher sensitivity to bold oxygenation level dependent (BOLD) signals and the ability to distinguish neurally related signals, multi-echo fMRI (ME-fMRI) has been a hot topic of research recently. Weighted combination in ME-fMRI is critical for the quality of the combined signal, which can be used for the analysis of brain functions. However, a convincing weighting method has never been put forward in previously published research. In this work, we optimized the weighting methods based on the Rayleigh quotient when using temporal signal-noise-ratio (tSNR) and multi-echo temporal signal-noise-ratio (metSNR) as metrics. These two metrics strongly represent the contrast-noise-ratio (CNR) which has been widely used to estimate fMRI quality. In our case of steady-state fMRI, with the optimal weighting methods, both tSNR and metSNR of the combined signal were improved significantly as compared to other widely-used weighting methods. In the meantime, we obtained negative weights that were not used in the combination for previous research, which drove us to find the causes and re-define the normalization as 'L1 Normalization'. We also found the distribution characteristics of negative weights in terms of brain region and TE. The occurrence and features of negative weights were explained by the correlation between echoes' data and the change of the multipliers (e.g. the mean value). In addition, we found that the metSNR performs more robustly than tSNR when comparing non-optimal methods to optimal methods and explained their variations based on the ratio from a mathematical viewpoint.