Integrating nonlinear optical and Raman spectral imaging for label-free pathological examinations

Nonlinear optical imaging (NLOI) provided detailed morphological information about biological systems, whereas confocal Raman micro-spectral imaging (CRMI) identified the biochemical properties of tissue samples. In this work, we proposed an integrated microscopy system by combining NLOI and CRMI together. An Er³⁺-doped femtosecond fiber laser at 1560 nm serves as the excitation source for NLOI modalities, and a semiconductor laser at 830 nm was used for spectra excitation during CRMI investigations. By taking a sectioned breast tissue sample to test the system's performance, three NLOI modalities for tissue characterization were realized, in which second harmonic generation (SHG) revealed collagen and elastin microstructures, third harmonic generation (THG) distinguished normal vs. diseased tissue architectures, and three-photon excited auto-fluorescence (3PEF) showed metabolic states of cells. Meanwhile, univariate Raman imaging illustrated the spatial distribution of important biochemical components, such as tryptophan (760 cm^-1), collagen (864 cm^-1, 945 cm^-1), and nucleic acids (1585 cm^-1). K-means cluster analysis (KCA) was further applied to resolve morphological and chemical features into distinct spectral clusters. The resolutions of SHG, THG and 3PEF were achieved by 2.67 μm, 1.5 μm, and 1.67 μm respectively; while the lateral resolution of Raman is 1.27 μm. These results indicated that our system could acquire multiple types of information while maintaining satisfactory performance. Moreover, the system is applicable to other pathological samples as its label-free nature and multimodal contrast enable quantitative analysis across various tissue types, which highlighted its promise for future biomedical applications.