Widefield fluorescence correlation spectroscopy

The focus of this thesis the use of image series data as spatially-separated time series data. The utilization of correlation techniques to analyze this time series data yields important results regarding the systems in question. A large portion of this work relies on microfluidic devices as model systems for microscopic motion to develop the analysis methods presented.Through the use of pixel-pair cross correlation mapping the flow in a three-dimensional volume in a microfluidic device was measured. When used in conjunction with a Nipkow disk confocal scanner this method allows for the rapid acquisition and analysis of flow speeds in an entire volume of interest within a microfluidic device. Simulations indicate that the correlation analysis technique is more tolerant to a high-noise environment than comparable particle tracking procedures, allowing data from images with S/N values as low as 1.4 to yield accurate flow maps with sufficient data length. This result indicates a compatibility with smaller, dimmer fluorescent probes for flow mapping and to that effect fluorescently-labelled single antibodies have been used to map flow in a microfluidic channel. For results presented here flow speeds of hundreds of micrometers per second are presented but with appropriate experimental conditions flow speeds of tens of millimeters per second are compatible with this method.Extending this technique to operate on multiple pixels surrounding a central origin allows for the fitting of a full flow vector from image series data. The correlation amplitude between a pair of pixels follows a sin 2 dependence on the angle between the flow vector and spacing vector connecting the two pixels. By fitting an experimentally-determined correlation amplitude map for a subregion of the area of interest the flow vector in that local area can be accurately determined. This method has been successfully applied to mapping flow around a partial obstruction and in the entrance of a microfluidic channel. Experimental and simulated data demonstrate that the full flow vector can be accurately computed for an arbitrary direction with speeds up to a few hundred micrometers per second.The correlation amplitude maps have been utilized to determine the axial, out-of-plane flow angle as well. This axial-angle fitting method relies on the same Nipkow spinning disk confocal scanner and additional a piezo nosepiece capable of rapidly switching between multiple focal planes. As a result synchronous time-series data from multiple focal planes can be used to produce correlation amplitude maps that express both the planar and axial motion of flow. Single-plane techniques are able to only fit the in-plane projection of the three-dimensional flow vector while the method presented here allows for the axial contribution to be quantified as well. This method has been shown to accurately measure axial flow angles in simulated data as well as map the flow vectors in microfluidic devices showing large amounts of axial motion.Spatially-separated time series data central to the flow mapping techniques can also be used to generate images in other experimental modalities. For optically-modulatable fluorophores, mapping the power of the modulation frequency yields a demodulated image with selective enhancement of the modulatable fluorophore over an obscuring constant background. A synthetic route to generate polystyrene beads labelled with Rose Bengal, a modulatable organic dye, for use as a modulatable probe is presented here. In addition, the use of higher-order statistics on time-series data of stochastic fluorophore blinking can be used to generate super-resolution images. This technique has been extended to demonstrate its capacity for super-resolution images beyond relying on quantum dots and allowing imaging with emitters that blink on timescales faster than can be obtained with CCD-based acquisitions. As such, the use of higher-order statistical analysis on confocal image data utilizing the microsecond blinking timescales of DNA:Ag nanoclusters to generate super-resolution images has been experimentally demonstrated.In order to improve the quality of imaging there have been several novel noble-metal based fluorophores synthesized and characterized. Light-activated emissive silver nanoparticles templated with synthetic tyrosine-containing peptides have been presented. These particles are extremely bright and show enhanced Raman signals at the level of single diffusing particles. The use of tyrosine derivatives as reducing agents for silver clusters has been further investigated, resulting in several new fluorophores of sizes comparable to organic dyes. Finally, orange-emitting gold-glutathione particles of sizes less than 5 nm have been synthesized.As microscopy techniques become more advanced, the subsequent techniques in data analysis and fluorescent probes must also continue to improve. Here has been presented novel methods for analyzing image stacks as a collection of spatially-distributed time series data in conjunction with correlations to yield a wide variety of results. These methods can be easily adapted to operate on a wide range of experimental systems, including live cells or whole organisms. With the increased used of such advanced image analysis techniques it is hoped that long-held secrets in microscopic phenomena can be revealed.