Multi-spectral lifetime imaging: methods and applications

Abstract

The aim of this PhD project is to further develop multispectral life time imaging hardware and analyses methods. The hardware system, Lambda-Tau, generates a considerable amount of data at high speed. To fully exploit the power of this new hardware, fast and reliable data analyses methods are required. Here, a phasor approach is used in all analyses; this provides a fast, visual solution for analyzing the recorded data. After successful demonstration of the phasor approach for time gated lifetime images, and showing that the time gated phasor analyses is compatible with a broad range of gate configurations, we used a similar approach to create the spectral phasor method which found immediate application in our group. Chapter 2 describes the theoretical framework of a generalized phasor approach for the analyses of time resolved data. It takes into account finite sampling and truncation of decay curves. An analytical expression for the phasor reference semicircle is presented which is compatible with different gate configurations and widths of the measurement windows. A global solution is provided to extract two lifetimes and fractional intensities in a binary mixture. This method is applied to analyze images recorded with a time gated detection system with 4 time gates (LIMO). An error analysis is performed to evaluate the required number of photons in a global analysis for estimation of two lifetimes in the image. The generalized phasor approach is exemplified in a FRET experiment. Chapter 3 describes a novel technique for spectral unmixing based on a phasor approach. The spectral phasor approach is a fast, reliable and simple method to represent the hyper spectral images in a 2D plot and provides a solution for semi-blind unmixing. We show that for a 3 component system all different mixtures fall inside a triangle with vertices made up by the spectral phasors of the pure components. In this chapter we successfully demonstrate the unmixing of images of fluorescently labeled cells and of autofluorescent grass blades. We show compatibility of the spectral phasor method with different spectral imaging systems; the spectral phasor approach can be employed even in systems with low numbers of spectral channels. In chapter 4 spectral phasor analysis is used to analyze two photon excited autofluorescence and second harmonic generation images of in vivo human skin. Different structures, including Stratum Corneum, epidermal cells, melanized cells and dermis, can be discerned by applying spectral phasor analyses. In chapter 5 the hardware development of the Lambda-Tau system is described. The hardware is based on a Field-Programmable Gate Array (FPGA), is capable of operating at high photon count rates and provides good time resolution. This chapter also contains an error analysis to investigate the performance of time domain lifetime imaging and spectral detection using different time gate and spectral channel configurations. Good performance of the electronics was demonstrated at count rates as high as 12 MHz per spectral detection channel. Chapter 6 is focused on the application of spectrally resolved lifetime imaging for blind unmixing. Time domain and spectral phasors are used for unmixing of the fluorescent components. Fractional intensities, spectra and decay curves of components are extracted without any prior knowledge. Chapter 7 provides a prospect to analyze the FRET samples using the information from all spectral channels. A model is provided based on phasor approach to simplify the behavior of ingrowth in acceptor channel. We test the method with fluorescently labeled DNA and we demonstrate the possibility of extraction FRET efficiency in presence of non-interacting and interacting donor molecules