Computational and Experimental Methods for Super-Resolution Imaging in Lensfree Microscopy

Abstract

Conventional light microscopes lack the ability to produce high-resolution, large field-of-view images in an inexpensive and timely manner. Although light microscopy is a common method for imaging pathology slides for disease detection and diagnosis, the capital and time constraints limit the ability to distribute these devices in point-of-care or low-resource settings as well as the ability to perform rapid disease diagnosis. Lensfree microscopes have emerged as economical and portable devices capable of producing images with high-resolution over a large field-of-view. Lensfree microscopes utilize few components, needing only a light source, sample to image, a sensor to record the light scattered from the illuminated object and a computer to perform the imaging or the object field reconstruction. Capable of producing amplitude and phase images, 3D volume images, and color images, the resolution of lensfree microscopes can be pushed further to expand the number of applications. In this dissertation I discuss computational and experimental methods for continuing to push the resolution and sensitivity of these systems beyond the micron scale. In the second chapter, I discuss computational methods for improving the speed and accuracy of image reconstructions for lensfree digital holography on the nanoscale by proposing nanoscale scalar scattering models. In the following chapter, I discuss a method for improving the detection of airborne particulate matter pollutants in lensfree digital holographic systems by increasing the scattering cross-section of sub-pixel sized particles using a pre-deposited film. Finally, I discuss the first implementation of time-gated fluorescence lensfree imaging for improving the SNR of lensfree fluorescence imaging, removing the need for spectral filters between the sample and sensor.