DNA is packaged into the nucleus of a mammalian cell as a nucleoprotein complex called chromatin. Changes in chromatin structure occur during processes that are critical to an understanding of mammalian cell biology such as cell division. Existing fixed-cell techniques have provided insight into chromatin organization in the mammalian nucleus. In addition, fluorescence microscopy techniques have allowed us to study changes in chromatin structure in living cells. However, most of these fluorescence techniques cannot be used for tissue imaging or long-term imaging due to photobleaching. In this thesis, we demonstrate that a label-free technique called Stimulated Raman Scattering (SRS) microscopy can be used to solve these problems and study chromatin structure in living mammalian cells both in culture and in tissue. SRS is a vibrational microscopy technique that takes advantage of intrinsic contrast arising from specific chemical bonds in a molecule. Nucleic acids have specifc phosphate and CH vibrations that can be used to determine their cellular distributions. Imaging at specific phosphate peaks using fingerprint SRS microscopy allows the detection of polytene chromosomes in Drosophila salivary gland cells and condensed chromatin in metaphase mammalian cells. In addition, we develop a technique called multicolor SRS microscopy, in which we image at several wavelengths across the CH vibrational band, and then use linear combination to simultaneously determine the nucleic acid, lipid and protein distributions in living mammalian cells. This technique achieves greater contrast than imaging at the phosphate vibrational peak due to the stronger SRS signal in the high wavenumber CH band and so allows us to determine chromatin structure in interphase mammalian cells. This technique also allows long-term imaging of living mammalian cells and the imaging of tissue such as mouse skin. The technique is used to monitor mammalian cell division in culture and paves the way for similar studies in living tissue. This technique will provide insight into cell division, differentiation and apoptosis during development and in disease models such as cancer.

Label-free imaging modalities offer numerous advantages, such as the ability to avoid the time-consuming and potentially disruptive process of physical staining. However, one challenge that arises in label-free imaging is the limited ability to extract specific structural or molecular information from the acquired images. To overcome this limitation, a novel approach known as digital staining or digital labeling has emerged. Digital staining leverages the power of deep learning algorithms to virtually introduce labels or stains into label-free images, thereby enabling the extraction of detailed information that would typically require physical staining. The integration of digital staining with label-free imaging holds great promise in expanding the capabilities of imaging techniques, facilitating improved analysis, and advancing our understanding of biological systems at both the cellular and tissue level. In this thesis, I explore supervised and semi-supervised methodologies of digital staining and the applications in augmenting label-free imaging modalities, particularly in the context of cell imaging and brain imaging. In the first part of the thesis, I demonstrate the novel integration of multi-contrast dark-field reflectance microscopy and supervised deep learning to enable subcellular immunofluorescence labeling and cell cytometry from label-free imaging. By leveraging the rich structural information and sensitivity of reflectance microscopy, this method accurately predicts subcellular features without the need for physical staining. As a result of the use of a novel multi-contrast modality, the digital labeling approach demonstrates significant improvements over the state-of-the-art techniques, achieving up to 3× prediction accuracy. In addition to fluorescence prediction, the method successfully reproduces single-cell level structural phenotypes related to cell cycles. The multiplexed readouts obtained through digital labeling enable accurate multi-parametric single-cell profiling across a large cell population. In the second part, I investigated a novel digital staining optical coherence tomography (DS-OCT) modality combining advantages of serial sectioning OCT and semi-supervised deep learning and demonstrated several advantages for the application of 3D histological brain imaging. The DS model is trained using a semi-supervised learning framework that incorporates unpaired translation, a biophysical model, and cross-modality image registration, which manifests broad applicability to other weakly-paired bioimaging modalities. The DS model enables the translation of S-OCT images to Gallyas silver staining, providing consistent staining quality across different samples. I further show that DS enhances contrast across cortical layer boundaries and enables reliable cortical layer differentiation. Additionally, DS-OCT preserves 3D-geometry on centimeter-scale brain tissue blocks. My pilot study demonstrates promising results on other anatomical regions acquired from different S-OCT systems, highlighting its potential for generalization in various imaging contexts. Overall, I investigate the problems of augmenting label-free imaging modalities with deep learning generated digital stains. I explored both supervised and semi-supervised methods for building novel DS frameworks. My work showcased two important applications in the field of immunofluorescence cell imaging and 3D histological brain imaging. On the one hand, the integration of DS techniques with multi-contrast microscopy has the potential to enhance the throughput of single-cell imaging cytometry, and phenotyping. On the other hand, integrating DS techniques with S-OCT holds great potential for high-throughput human brain imaging, enabling comprehensive studies on the structure and function of the brain. Through the exploration, I aim to shed light on the impact of digital staining in the field of computational imaging and its implications for various scientific disciplines.

Biomedical research relies on identification and isolation of specific cell types using molecular biomarkers and sorting methods such as fluorescence or magnetic activated cell sorting. Labelling processes potentially alter the cells’ properties and should be avoided, especially when purifying cells for clinical applications. A promising alternative is the label-free identification of cells based on physical properties. Sorting real-time deformability cytometry (soRT-DC) is a microfluidic technique for label-free analysis and sorting of single cells. In soRT-FDC, bright-field images of cells are analyzed by a deep neural net (DNN) to obtain a sorting decision, but sorting was so far only demonstrated for blood cells which show clear morphological differences and are naturally in suspension. Most cells, however, grow in tissues, requiring dissociation before cell sorting which is associated with challenges including changes in morphology, or presence of aggregates. Here, we introduce methods to improve robustness of analysis and sorting of single cells from nervous tissue and provide DNNs which can distinguish visually similar cells. We employ the DNN for image-based sorting to enrich photoreceptor cells from dissociated retina for transplantation into the mouse eye.