Targeted therapy is crucial to improving the prognosis of gastric cancer and thus decreasing its mortality. The premise of designing targeted therapy strategy is the identification and subtype classification of malignant gastric lesions. Spectrum-resolved multiphoton microscopy (SMPM) is capable of providing not only structural, but also biochemical information at subcellular level. Comparing with conventional techniques that generally focus on qualitative morphological characterizing of gastric mucosa, this multidimensional imaging technique may provide more powerful diagnostic capabilities. However, its full potential value has not been extensively evaluated in clinical settings. Here, we performed an investigation on human gastric cancers based on label-free SMPM imaging of fresh tissue specimens of normal gastric mucosa, intestinal-type adenocarcinoma, and neuroendocrine carcinoma. By extracting emission spectral information of endogenous fluorophores, the three-dimensional subcellular histology and biochemical components of gastric mucosa are revealed. Based on these clues, the sub-structures of gastric mucosa, including surface epithelium, interstitial tissue of lamina propria, and gastric pit, are clearly identified. Furthermore, qualitative and quantitative indicators based on the SMPM signals of gastric mucosa were created, which are found to have the potential to discriminate normal gastric mucosa and different types of gastric cancers. This study fills the knowledge gap of human malignant gastric lesions under SMPM imaging and may shed new light on the diagnosis and classification of gastric cancers. With advances in multiphoton endoscopy, the SMPM has the potential to be developed into a noninvasive, label-free, real-time histological and functional diagnosis instrument in the future.
Many imaging technologies, such as two photon microscopy (TPM), second generation microscopy (SHG) and photoacoustic microscopy (PAM), have successfully demonstrated the ability to extract anatomical, functional and molecular information of biological samples. However, they always fail to obtain comprehensive information from biological tissue due to single imaging modality. To address this limitation, we developed a multimodal microscopy with fully integrated PAM, TPM and SHG. The home-built multimodal microscopy system enable label-free imaging of biomedical tissue with sub-micron resolution in vivo and in vitro. The results may offer a new tool to provide the capable of fast comprehensive information capturing for biological studies.