Innovations in Imaging System Design: Gigapixel, Chip-Scale and MultiFunctional Microscopy

Abstract

Microscopy imaging is of fundamental importance in diverse disciplines of science and technology. In a typical microscopy imaging platform, the light path can be generalized to the following steps: photons leave the light source, interact with the sample, and finally are detected by the image sensor. Based on such a light path, this thesis presents several new microscopy imaging techniques from three aspects: illumination design, sample manipulation, and imager modification. The first design strategy involves the active control of the illumination sources. Based on this strategy, we demonstrate a simple and cost-effective imaging method, termed Non-interferometric Aperture-synthesizing Microscopy (NAM), for breaking the spatial-bandwidth product barrier of a conventional microscope. We show that the NAM method is capable of providing two orders of magnitude higher throughput for most existing bright-field microscopes without involving any mechanical scanning. Based on NAM, we report the implementation of a 1.6 gigapixel microscope with a maximum numerical aperture of 0.5, a field-of-view of 120 mm2, and a resolution-invariant imaging depth of 0.3 mm. This platform is fast (acquisition time of ~ 3 minutes), free from chromatic aberration, capable for phase imaging, and, most importantly, compatible with most existing microscopes. High quality color images of histology slides were acquired by using such a platform for demonstration. The proposed NAM method provides a robust way to transform the problem of high-throughput microscopy from one that is tied to physical limitations of the optics to one that is computationally solvable. The active control of illumination sources can also be adapted for chip-scale microscopy imaging. To this end, we present a lensless microscopy solution termed ePetri-dish. This ePetri-dish platform can automatically perform high resolution (~ 0.66 micron) microscopy imaging over a large field-of-view (6 mm × 4 mm). This new approach is fully capable of working with cell cultures or any samples in which cells/bacteria may be contiguously connected, and thus, it can significantly improve Petri-dish-based cell/bacteria culture experiments. With this approach providing a low-cost and disposable microscopy solution, we can start to transit Petri-dish-based experiments from the traditionally labor-intensive process to an automated and streamlined process. The second strategy in design considerations is to manipulate the sample. We present a fully on-chip, lensless, sub-pixel resolving optofluidic microscope (SROFM). This device utilizes microfluidic flow to deliver specimens directly across an image sensor to generate a sequence of low-resolution projection images, where resolution is limited by the sensor’s pixel size. This image sequence is then processed to reconstruct a single high-resolution image, where features beyond the Nyquist rate of the LR images are resolved. We demonstrate the device’s capabilities by imaging microspheres, protist Euglena gracilis, and Entamoeba invadens cysts with sub-cellular resolution. The third accessing point in design considerations is the image sensor. Imager modification is an emerging technique that performs pre-detection light field manipulation. We present two novel optical structure designs: surface-wave-enabled darkfield aperture (SWEDA) and light field sensor. These structures can be directly incorporated onto optical sensors to accomplish pre-detection background suppression and wavefront sensing. We further demonstrate SWEDA’s ability to boost the detection sensitivity, with a contrast enhancement of 27 dB.