Single-cell feature analysis method and device based on hyperspectral coherent raman scattering imaging, equipment and medium

Abstract

This invention relates to a method, apparatus, device, and medium for single-cell feature analysis based on hyperspectral coherent Raman scattering imaging, used for single-cell feature analysis. The invention provides an apparatus for performing hyperspectral coherent Raman scattering imaging of samples at different Raman shifts; the method performs hyperspectral coherent Raman imaging on samples enriched with cells using medium A, acquiring images and Raman spectra of different substances at multiple Raman shifts; a multivariate curvature resolution method is used to obtain concentration distribution maps of different substances, and an image is obtained after removing medium A; deep learning technology is used to segment single-cell regions, extracting morphological and metabolic features of cells; simultaneously, a spectral phasor method combined with Lorentz fitting is used to extract morphological and metabolic features of single cells and their subcellular structures. This invention solves the problem of medium structure occlusion, resulting in richer extracted morphological and metabolic features of cells, facilitating more accurate differentiation of cell types.

Images