Label-free identification of early gastrointestinal neuroendocrine tumors via biomedical multiphoton microscopy and automatic image analysis.

At present, early diagnosis and treatment is the most effective way to treat early gastrointestinal neuroendocrine tumors. Therefore, we attempted to carry out multiphoton imaging of early neuroendocrine tumors because of its ability to label-free image tissue microstructure at the cellular level. Imaging results show that this imaging technique can quickly identify the histopathological changes in mucosa and submucosa caused by tumor invasion. Furthermore, we performed automatic image analysis on SHG images and extracted two optical diagnostic features-collagen density and average intensity, and also found obvious differences in the density as well as average intensity of collagen fibers in tumor microenvironment using a series of quantitative analysis. These findings may further facilitate the development of multiphoton microscopic imaging technique for clinical use.