Abstract
Understanding cells as integrated systems is a challenge central to modern biology. While different microscopy approaches may be used to probe diverse aspects of biological organization, each method presents limitations which ultimately restrict a view into unified cellular organization. For example, while fluorescence microscopy can resolve subcellular structure in living cells, it is expensive, slow, and can damage cells. Here, we present a label-free method for predicting 3D fluorescence directly from transmitted light images and demonstrate that it can be used to generate multi-structure, integrated images. We then demonstrate that this same method can be used to predict immunofluorescence from electron micrograph inputs, extending the method to a wider range of bioimaging applications.
