Study Of The Constructions And Applications Of Intelligent Diagnostic Model Of Esophageal Squamous Cell Carcinoma Based On Tissue Microarrays Resource

Objective:To construct a platform of tissue microarrays resources and illustrate its value in the tumor progression research by setting the progression-based series tissue microarrays of esophageal squamous cell carcinoma as an example.Tissue microarrays resources,combined with image analysis,data mining and artificial neural network,were used to establish a pathological diagnosis model to explore new ideas of application of tissue microarrays and intelligent diagnosis for clinical pathology.Methods:In order to construct of tissue microarrays resources platform,a large number of stored paraffin block were made into tissue microarrays specimen and the corresponding data management system database was established by ACCESS.The expressions of the key proteins of Wnt pathway were detected by immunohistochemistry using progression based series tissue microarrays of ESCC. Then lots of experimental photographs were analysis by image analysis.Using statistical analysis to determine the best cut-off point of optical density between normal and cancerous tissues,then intelligent diagnosis model of ESCC based on tissue microarrays resources was established by neural network technology. Subsequently,we trained and simulated the model by data sets and evaluated the diagnostic results and mined the optimal combination of indicators for diagnosis finally.Results:At present,tissue microarrays resources stored 35 tissue microarrays paraffin blocks,the specimens consisting of esophageal squamous cell carcinoma,colorectal cancer,gastric cancer,the total number of cases were more than 1100,the total number of sections were more than 700 and some of them had played roles in the clinical and basic research;Simultaneously the tissue microarrays data management system design and entry work have been initially completed.Wnt pathway may play an important role in the tumorigenesis of esophageal squamous cell cancer based on tissue microarrays and immunohistochemical results.Image analysis results showed that normal and cancerous esophageal tissues were significantly differences,the diagnosis results of artificial neural network model established on their optical density cut-off point showed that single indicator could not accurately distinguish them and the accuracy of combination of indicators were better than the single indicator.Conclusion:The tissue microarrays have so widespread use that the clinical and scientific experiments will be carried out greatly facilitate for the tissue microarrays resources.Intelligent diagnosis provides a new idea for the application of tissue microarrays;artificial neural network technology will improve the diagnostic effectually.