New implementations of phase-contrast imaging

Phase-contrast imaging is a method of imaging widely used in biomedical research and applications. It is a label-free method that exploits intrinsic differences in the refractive index of different tissues to differentiate between biological structures under analysis. The basic principle of phase-contrast imaging has inspired a lot of implementations that are suited for different applications. This thesis explores multiple novel implementations of phase-contrast imaging in the following order. 1, We combined scanning Oblique Back-illumination Microscope (sOBM) and confocal microscope to produce phase and fluorescence contrast images in an endomicroscopy configuration. This dual-modality design provides co-registered, complementary labeled and unlabeled contrast of the sample. We further miniaturized the probe by dispensing the two optical fibers in our old design. And we presented proof of principle demonstrations with ex-vivo mouse colon tissue. 2, Then we explored sOBM-based phase and amplitude contrast imaging under different wavelengths. Hyperspectral imaging is achieved by multiplexing a wide-range supercontinuum laser with a Michaelson interferometer (similar to Fourier transform spectroscopy). It features simultaneous acquisition of hyperspectral phase and amplitude images with arbitrarily thick scattering biological samples. Proof-of-principle demonstrations are presented with chorioallantoic membrane of a chick embryo, illustrating the possibility of high-resolution hemodynamics imaging in thick tissue. 3, We focused on increasing the throughput of flow cytometry with principle of phase-contrast imaging and compressive sensing. By utilizing the linearity of scattered patterns under partially coherent illumination, our cytometer can detect multiple objects in the same field of view. By utilizing an optimized matched filter on pupil plane, it also provides increased information capacity of each measurement without sacrificing speed. We demonstrated a throughput of over 10,000 particles/s with accuracy over 91% in our results. 4, A fourth part, which describes the principle and preliminary results of a computational fluorescence endomicroscope is also included. It uses a numerical method to achieve sectioning effect and renders a pseudo-3D image stack with a single shot. The results are compared with true-3D image stack acquired with a confocal microscope