Phase imaging with computational specificity for cell biology applications

Abstract

Recent advancements in quantitative phase imaging (QPI) and deep learning have opened up an exciting frontier. It has been shown that deep learning methods, with their ability to extract intricate structures from massive raw datasets, can be applied to both interpreting QPI measurements of biological samples and enhancing the imaging capabilities of QPI systems. Phase imaging with computational specificity (PICS), a workflow that combines deep learning and QPI, has recently been developed to nondestructively measure biophysical parameters or markers from label-free samples directly. In this thesis, we present a new non-invasive, high-throughput method built upon the principle of PICS, to detect the cell cycle of live cell clusters. We demonstrate that the proposed method can be applied to study single-cell dynamics within the cell cycle as well as investigate cell biophysical parameter distribution across different stages of the cell cycle.