Quantitative three dimensional maps of cellular structure, activity and function provide the key to answering many prevalent questions in modern biological research. Fluorescence microscopy has emerged as an indispensable tool in generating such maps, but common techniques are limited by fundamental physical constraints which render them incapable of simultaneously achieving high spatial and temporal resolution. This thesis will describe the development of novel microscopy techniques and complementary computational tools capable of addressing some of the aforementioned limitations of fluorescence microscopy and further outline their application in providing novel biological insights. The first section details the design of a light sheet microscope capable of high-throughput imaging of cleared, macroscopic samples with cellular resolution. In light sheet microscopy, the combination of spatially confined illumination with widefield detection enables multi-megapixel acquisition in a single camera exposure. The corresponding increase in acquisition speed enables systems level biological studies to be performed. The ability of this microscope to perform rapid, high-resolution imaging of intact samples is demonstrated by its application in a project which established a niche and hierarchy for stem cells in the adult nervous system. Light sheet microscopy achieves fast volumetric imaging rates, but the two dimensional nature of each measurement results in an inevitable lag between acquisition of the initial and final planes. The second section of this thesis describes the development and optimization of a light field microscope which captures volumetric information in a snapshot. Light field microscopy is a computational technique and images are reconstructed from raw data. Both the fidelity of computed volumes and the efficiency of the algorithms are strongly dependent on the quality of the rectification. A highly accurate, automated procedure is presented in this section. Light field reconstruction techniques are investigated and compared and the results are used to inform the re-design of the microscope. The new optical configuration is demonstrated to minimize the long-object problem. In the final section of the thesis, the spatial resolution limits of light field microscopy are explored using a combination of simulations and experiments. It is shown that light field microscopy is capable of localizing point sources over a large depth of field with high axial and lateral precision. Notably, this work paves the way towards frame rate limited super resolution localization microscopy with a depth of field larger than the thickness of a typical mammalian cell.

The aim of the work described in this thesis is to improve the understanding, implementation, and overall capabilities of structured illumination microscopy (SIM). SIM is a superresolution technique that excels in gentle live-cell volumetric imaging tasks. Many modalities of SIM were developed over the last decade that tailored SIM into the versatile and powerful technique that it is today. Nevertheless, the field of SIM continues to evolve and there is plenty of room for novel concepts. Specifically, in this thesis, a generalised framework for a theoretical description of SIM variants is introduced, the constraints of optical components for a flexible SIM system are discussed and the set-up is realised, the important aspect of deconvolution in SIM is highlighted and further developed, and finally novel SIM modalities introduced that improve its time-resolution, gentleness, and volumetric imaging capabilities. Based on the generalised theory, the computational steps for the extraction of superresolution information from SIM raw data are outlined and the essential concept of spatial frequency un-mixing explained for standard SIM as well as for multifocal SIM. Multifocal SIM hereby acts as a parallelised confocal as well as widefield technique and thus serves as link between the two modalities. Using this novel scheme deconvolution methods for SIM are then further developed to allow a holistic reconstruction procedure. Deconvolution is of great important in the SIM reconstruction process, and hence rigorous derivations of advanced deconvolution methods are provided and further developed to enable generalised ‘multi-image’ Richardson-Lucy deconvolution in SIM, called joint Richardson-Lucy deconvolution (jRL). This approach is demonstrated to robustly produce optically sectioned multifocal SIM images and, through the incorporation of a 3D imaging model, also volumetric standard SIM images within the jRL framework. For standard SIM this approach enabled acquisition speed doubling, because the recovery of superresolved images from a reduced number of raw frames through constrained jRL was made possible. The method is validated in silico and in vitro. For the study of yet faster moving samples deconvolution microscopy is found to be the method of choice. To enable optical sectioning, a key feature of SIM, in deconvolution microscopy, a new modality of optical sectioning microscopy is introduced that can be implemented as a single-shot technique. Via polarised excitation and detection in orthogonal directions in conjunction with structured illumination the theoretical framework is rigorously derived and validated.

Volumetric optical microscopy has the advantages of quantitative and global measurement of three-dimensional (3D) biological specimens with high spatial resolution and minimum invasion. However, current volumetric imaging technologies based on light transmission, scattering or fluorescence cannot reveal specimen’s chemical distribution that brings insights to study the chemical events in organisms and their metabolism, functionality, and development. Stimulated Raman scattering (SRS) microscopy allowing visualization of chemical contents based on their intrinsic molecular vibrations is an emerging imaging technology to provide rapid label-free volumetric chemical imaging. This dissertation describes three methodologies for developing advanced volumetric SRS imaging technologies to address the challenges of imaging in vivo samples, imaging speed, and axial resolution. In the first methodology, SRS volumetric imaging is enabled by axially scanning the laser foci for sectioning different depth layers. In Chapter 2, we utilize a piezo objective positioner to drive the objective. Combining with the tissue clearance technique, we realize volumetric SRS imaging up to 500 µm depth in brain tissues showing the potential for 3D staining-free histology. The limitations of piezo scanning are slow speed and disturbance to in vivo samples while rapidly scanning the objective. To tackle the limitations, in Chapter 3, we develop a remote-focusing volumetric SRS microscope based on a deformable mirror and adaptive optics optimization, allowing focal scanning without physically moving the objective or sample. We demonstrate in vivo monitoring of chemical penetration in human sweat pores. In the second methodology, instead of axially scanning the laser foci, the SRS volumetric imaging is enabled by projection imaging with extended depth-of-focus (DOF) beams such as Bessel beams and low numerical-aperture beams. The extended DOF beams integrate SRS signals along the propagation direction to form projection images; thus, a single lateral scan obtains the volumetric chemical information, significantly increasing the volumetric imaging speed for measuring chemical content over a large volume. In Chapter 4, we describe a stimulated Raman projection microscope for fast quantitation of chemicals in a 3D volume. However, projection imaging intrinsically loses axial resolution. We addressed the limitation by developing SRS projection tomography. Mimicking computed tomography, the axial information is reconstructed by angle-dependent projection images obtained by sequentially rotating the sample in a capillary glass tube within the SRS focus. Nevertheless, sample rotation is complicated and not compatible with in vivo samples. To address the difficulty, in Chapter 5, we develop tilted-angle-illuminated stimulated Raman projection tomography which utilizes tilted-angle beams with a tilted angle respected to the optical axis of the objective to obtain angle-dependent projections. This scheme is free of sample rotation and enables fast projection scanning for pushing the imaging speed. The calibration approach and vector-field back-projection algorithm are developed for the multi-view tomographic reconstruction. In the third methodology, we improve the spatial resolution in miniature volumetric SRS imaging via the innovation of metasurface photonics. In developing an SRS endoscope for volumetric chemical imaging inside the human body, the axial resolution deteriorates due to chromatic and monochromatic aberrations induced by poorly made miniature objective lenses. In Chapter 6, we develop a silicon metasurface tailored for compensating the phase errors between the pump and Stokes wavelengths of a singlet refractive lens. Integrating the metasurface with the refractive lens, the hybrid achromatic metalens is compact and provides nearly diffraction-limit resolution, demonstrating a way for developing high resolution chemical imaging endoscopy.