Label-free microscopy is a promising imaging platform in biophysics. It preserves physiology, minimizes preparation bias, and provides highly quantitative imaging. In parallel, virtual staining (also called digital staining, in silico labeling, or label-free prediction), which predicts stained images from label-free inputs, has been demonstrated in selected tissues, cells, and organelles. However, most prior work has focused on larger structures. Extending this paradigm to finer targets is an important goal in quantitative cell biophysics. In this study, we introduce UpDPC (upgraded polarization-resolved differential phase contrast), a computational phase modality that transforms standard commercial microscopes into single-shot, label-free nanoscopy platforms with a simple add-on. UpDPC retrieves quantitative phase contrast over a wide range of spatial frequencies from a single camera exposure. This enables motion-robust, low-phototoxic imaging in live cells. The static, single-exposure optical setup minimizes hardware controls and motion artifacts, facilitating the observation of delicate morphologies during fast cellular events. Using this system, we can track the dynamics of fine intracellular structures, such as mitochondrial membranes, endoplasmic reticulum (ER) networks, and actin stress fibers, in real time. We demonstrate the diversity of organelle morphologies accessible using UpDPC. Our results establish UpDPC as a practical tool for routine, label-free tracking of rapid intracellular dynamics on widely available hardware. We also briefly introduce recent virtual staining results driven by UpDPC inputs, targeting finer organelles, such as the ER, demonstrating its compatibility and future potential.
Inutsuka et al. (Sun,) studied this question.