Stimulated Raman scattering (SRS) microscopy is a powerful chemical imaging tool for visualizing biomolecule distribution in biological cells and tissues. Recent research has focused on SRS imaging in the C-H region due to the strong signals from lipids and proteins. However, these signals are regarded as non-specific. To improve the specificity of C-H imaging, we sought to use advanced machine learning to extract hidden information from C-H SRS imaging. This is possible because cells in tissue often have distinct sizes, shapes, and compositions. In this talk, I will present our recent efforts on machine learning/deep learning augmented cell imaging and classification with SRS. This approach potentially enables label-free mapping and tracking of different cells in various tissue.
|