Breaking through imaging resolution and depth barriers with diffuse optical localization methods

Abstract

Fluorescence microscopy has emerged as an indispensable tool for biological research, revealing intricate structural and functional details at cellular and sub-cellular levels. Despite significant advancements in this field, certain challenges persist, which impede high-speed large-scale deep tissue investigations with fluorescence microscopy. Specifically, spatial resolution is confined by the optical diffraction limit and further compromised by light scattering in turbid specimens. On the other hand, temporal resolution, crucial for capturing fast biodynamics, is constrained by the scanning or detection speeds of associated scanners and photodetectors. Furthermore, the pronounced effects of light diffusion and absorption curtail the effective penetration depth in biological tissues to ~1 mm within the ballistic regime of light. Surmounting these resolution and depth barriers constitutes a major challenge for fluorescence microscopy. To unlock the full potential of fluorescence microscopy, this dissertation delves into its two foundational design paradigms: widefield fluorescence microscopy (WFM) and laser scanning fluorescence microscopy (LSFM). Within the WFM framework, the diffuse optical localization imaging (DOLI) technique is introduced, which leverages the principle of localization and operates in the second near-infrared window to achieve superior resolution-penetration balance in scattering media. Transitioning to LSFM, the large-field multifocal illumination microscopy (LMI) and its derivative, the parallel-exposure LMI (PE-LMI), have been developed, aiming at substantially accelerating the imaging speed. Furthermore, both DOLI and LMI are adapted to enable three-dimensional imaging, grounded on stereovision and optical astigmatism, respectively. Lastly, the dissertation showcases applicability of the proposed microscopic methods in cerebrovascular imaging of murine brains under both healthy and pathological conditions, depicting the blood flow regulation of the entire pial vascular network at unprecedented detail. Collectively, the devised methods offer a wealth of possibilities for non- or minimally-invasive imaging of large-scale biodynamics, with an unmatched combination of high spatio-temporal resolution and large field of view.