Manuscript Summary - Unsupervised Hyperspectral Stimulated Raman Microscopy Image Enhancement: De-Noising and Segmentation via One-Shot Deep Learning
Raman microscopes let you observe how a sample responds to light across a wide range of energies. Different chemicals behave differently, so these responses can be used to identify different chemicals within a sample. Unfortunately, images can sometimes be noisy, and the identification process timeconsuming and labour intensive. Here we have used unsupervised machine learning to remove noise in the data, and find patterns within the sample - identifying regions which are made up material which behaves similarly. This is useful for producing chemical "maps", making the microscope much more intuitive and useful.