Rapid and label-free detection of gastrointestinal stromal tumor via a combination of two-photon microscopy and imaging analysis

Abstract

Background Gastrointestinal stromal tumor (GIST) is currently regarded as a potentially malignant tumor, and early diagnosis is the best way to improve its prognosis. Therefore, it will be meaningful to develop a new method for auxiliary diagnosis of this disease. Methods Here we try out a new means to detect GIST by combining two-photon imaging with automatic image processing strategy. Results Experimental results show that two-photon microscopy has the ability to label-freely identify the structural characteristics of GIST such as tumor cells, desmoplastic reaction, which are entirely different from those from gastric adenocarcinoma. Moreover, an image processing approach is used to extract eight collagen morphological features from tumor microenvironment and normal muscularis, and statistical analysis demonstrates that there are significant differences in three features—fiber area, density and cross-link density. The three morphological characteristics may be considered as optical imaging biomarkers to differentiate between normal and abnormal tissues. Conclusion With continued improvement and refinement of this technology, we believe that two-photon microscopy will be an efficient surveillance tool for GIST and lead to better management of this disease.

Detection of fibrotic changes in the progression of liver diseases by label‐free multiphoton imaging

“…For example, Chen et al used multiphoton imaging to study the collagen characteristics of gastric tumor microenvironment and successfully uncovered optical markers of lymph node metastasis [32]. Li et al used MPM to image gastrointestinal stromal tumor and found significant differences in collagen fibers between normal and cancerous regions [33]. In this study, we found that SHG imaging can be used to accurately identify the different stages of fibrosis and to monitor the spatial distribution of collagen fibers during disease progression.…”

Prognostic significance of collagen signatures in pancreatic ductal adenocarcinoma obtained from second-harmonic generation imaging

“…1 A). Then, we developed an automatic image processing method to quantitatively analyze collagen changes in SHG images [ 24 , 25 ]: a Gaussian mixture model-based segmentation algorithm was used to segment SHG images into collagen fibers and background; afterwards, to track each fiber, we used a mature fiber network extraction algorithm to process the binary mask image of collagen fibers. After fiber extraction, the skeleton of each fiber was identified and represented by a list of ordered vertices, and if any vertex in the list belongs to more than one fiber, it would be identified as a cross-link point.…”

Application of multiphoton microscopy for studying liver fibrosis

“…As described previously, we employed a commercial laser scanning confocal microscopy (LSM 880, Zeiss, Germany) in conjunction with a mode-locked femtosecond Ti: sapphire laser (Chameleon Ultra, Coherent, USA) to conduct multiphoton imaging [8][9][10]. A Plan-Apochromat 20×objective (NA = 0.8, Zeiss, Germany) was employed to acquire multiphoton images.…”