| Most biological cells are not clearly visible with a bright field microscope. Several methods have been developed to improve contrast in cell imaging, including use of exogenous contrast agents such as fluorescence microscopy, as well as utilizing properties of light-specimen interaction for optics design, to reveal the endogenous contrast, such as phase contrast microscopy (PCM) and differential interference contrast (DIC) microscopy. Although PCM and DIC methods significantly improve the image contrast without the need for staining agents, they only provide qualitative information about the phase change induced by the cells as light passes through them. Quantitative phase imaging (QPI) has recently emerged as an effective imaging tool which provides not only better image contrast but also cell-induced phase shifts in the optical pathlength, thus allowing nanometer-scale measurements of structures and dynamics of the cells. Other important aspects of an imaging system are its imaging speed and throughput. High-throughput, high-speed, real-time quantitative phase imaging with high spatial and temporal sensitivity is highly desirable in many applications including applied physics and biomedicine. In this dissertation, to address this need, I discuss the development of such an imaging system that includes the white light diffraction phase microscopy (wDPM), a new optical imaging method, and image reconstruction/analysis algorithms using graphics processing units (GPUs). wDPM can measure optical pathlength changes at nanometer scale both spatially and temporally with single-shot image acquisition, enabling very fast imaging. I also exploit the broadband spectrum of white light used as the light source in wDPM to develop a system called spectroscopic diffraction phase microscopy (sDPM). This sDPM system allows QPI measurements at several wavelengths, which solves the problem of thickness and refractive index coupling in the phase shifts induced by the cell, and which also may help visualize more-complex cell structures. Owing to its high spatial and temporal sensitivity and single-shot acquisition, wDPM enables measurement of nanometer-scale dynamic processes of cells at very high rate and measurement of cell growth because of the linear relationship between a cell-induced phase shift and its dry mass. The parallel algorithms and software tools I developed allow real-time QPI imaging and online image analysis at frame rates of up to 40 megapixel-size images per second. This capability allows very high throughput of several thousands of cells in imaging mode and eliminates the need of storing the images since we only need to store processed data, which is much smaller in storage size. Finally, I present the capability of the system by showing an application in red blood cell screening, which can be used as a diagnostic tool in blood testing and may pave the way for digital hematology and remote diagnostics. |
