New method could democratize deep learning-enhanced microscopy

Deep learning is a potential tool for scientists to glean more detail from low-resolution images in microscopy, but it’s often difficult to gather enough initial data to train computers in the process. A new method developed by Salk Assistant Research Professor Uri Manor, director of Salk’s Waitt Advanced Biophotonics Core Facility, and first author Linjing Fang, Salk image analysis specialist, could make the technology more accessible by taking high-resolution images and artificially degrading them. The method could make it significantly easier for scientists to get detailed images of cells that have previously been difficult to observe, as well as allow scientists to capture high-resolution images even if they don’t have access to powerful microscopes.