Digital Construction Of Both Precise Surveillance On Gastric Precancerous Lesions And Effective Reversion System Of Gastric Canceration

Gastric cancer is a common malignant tumor worldwide. The prognosis is strongly related with the early diagnosis. Until now, there is still lack of effective diagnostic and monitoring methods of gastric cancer. Many cases were diagnosed as gastric cancer until they have reached the end stage. The development and progression of most gastric cancer may involve a common route:persistent chronic gastritis, gastric glandular atrophy, intestinal metalplasia, dysplasia, and then canceration. Therefore, it is very important to develop reliable methods for accurate diagnosis and effective prevention/treatment of precancerous lesions. We have collected the information about the cancer-related clinical risk factors, pathological changes, expression of proteins and genes of the patients with precancerous lesions. Using the data to perform group- and individual-level analysis, prospective observational study and pilot survey on precancerous lesions, we work out the principles on gastric precancerous surveillance, early warning of canceration and treatment of gastric precancerous lesions. We construct a clinical platform to establish an effective monitoring and reversion/treatment system of gastric precancerous lesions. Practically, we are facing three obstacles:1.The accuracy of gastric endoscopic biopsies. Although we have invented a target marking biopsy technique (MTB:marker target biopsy, a technique with injecting India ink into the mucous that can mark the biopsy point) to ensure the accuracy of biopsy pathology, there are still several shortcomings:it is an invasive method with potential risks to induce cancerasion and other issues, which can lead to the limitation of the extensive use of this technology.2. Reliability of gastric canceration warning:This involves a comprehensive analysis of monitoring data about diagnosis and treatment of gastric cancer. Previous studies have established several digital formulas using genetic, microRNA, and protein information to predict carcinogenesis. But these studies were based on a limited population data from few laboratories, thus, a wide range of patient populations and a number of different laboratories should be recruited to get the actual range of these molecular markers. We aim to establish a high specificity and sensitivity digital model for the universal diagnosis of gastric cancer. Therefore, patients can be classified systemically by the digital model with automatic correction.3. Effective reversion/treatment of gastric cancer:This involves individual treatment. Several strategies had been launched for gastric pre-carcerous lesion reversion from previous interventions therapy, and also established measures on reducing injury, promoting tissue repair, anti-mutation and lesions resection. However, the overall effectiveness is still unsatisfied. There are too many factors involved and difficult to analyze. In fact, the current treatment strategies vary from one person to another and there are no uniform guidelines to evaluate the efficacy of the treatment. Thus, the accuracy and credibility of the treatment information from sufficient number of individuals should be collected before and after treatment, and should be analyzed by computer to find out the valid and invalid factors, and then set up a new effective program.The internet technology should be used to resolve these three issues. We need to establish an artificial intelligent (AI) system using information collected from clinical surveillance, endoscopic image, pathology, cellular, molecular genetics, other biological data and cancer reversion drugs of gastric precancerous lesions to help clinical diagnosis, treatment strategies, prevention and control of precancerous lesions. Such AI system will also be created on line for the patients to receive health education, simplify treatment procedure, reduce the workload of medical staff and greatly reduce medical costs.This study aims to establish the biological digital information platforms of precancerous lesions and AI system to help guiding the effective prevention and treatment of gastric cancer clinically. Basis on a previous work, our team has made improvements in our hospital and conduct research to resolve these three problems through a preliminary trial of precancerous lesions and cancer monitoring clinical intervention system. We now are focused on the establishment of an Internet-based system, the digital "accurate surveillance of precancerous lesions and effective reversion system of cancer diagnosis and treatment"Methods1. Based on the digital three-dimensional simulation re-construction of the stomach wall surface of the first endoscopy, the computer can re-localize the biospy point in the second endoscopy and navigate the endoscope to the biospy point in the first endoscopy. We call this new technique computer simulated marker target biopsy (CSMTB). Atrophic gastritis volunteers received the MTB as the marking standard. The accuracy of computer simulated marker target biopsy is detected during endoscopy.2. Use qRT-PCR method to test CST1 and HOXC6 level in gastric cancer, precancerous lesions and superficial gastitis, then perform discriminant analysis on collegan family.in 105 cases of gastric cancer, precancerous lesions and superficial gastitis. Finally, perform discriminant analysis on collegan family and CST1\HOXC6 and try to build up a discriminant model for gastric cancer.3. According to previous experiments required to collect gastric precancerous lesions, after gastroscopy with MTB, Misoprostol and Mucosta were given respectively together with folic acid for 6 months treatment, the clinical symptoms were recorded before and after the medication were given, we hope to propose a better reversal treatment system of gastric precancerous lesions initially combined with previous treatment regimen.Results1. We successfully constructed a non-invasive computer navigation system to guide the biopsy. Digital stomach linings live 3D dynamic model to simulate marking gastric biopsy is brought into the reality, easy to operate. Compared with the MTB, its accuracy is 7.23mm; The accuracy is high enough for clinic use.2. qQRT-PCR test results showed CST1 and HOXC6 expression levels were significantly higher in gastric cancer patients than in precancerous lesions and superficial gastritis patients (p<0.05). Together with collagen genes, gender and age, we established the discrimination model of gastric cancer, precancerous lesions and superficial gastritis. The discriminant accuracy is 95.8%.The patient’s age, COL4A6 \COL1A1\CST1\HOXC6 together were involved into the discriminant model by stepwise discriminant method, back substitution summary accuracy was 97.9% which can effectively distinguish gastric cancer, from precancerous lesions and superficial gastritis. Gastric cancer -related genes, RNA, proteins maybe combined together to help distinguish gastric cancer from precancerous lesions and superficial gastritis.3. MTB was used to mark gastric biopsy. The reversion rate of misoprostol with folic acid on atrophic gastritis and intestinal metaplasia was 70%, better than teprenone, rebamipide, and other non-specific treatment groups. Application of Internet technology to collect a large number of individuals,we analyzed the factors for the precancerous lesions reversion on bother the successful cases and failure cases and proposed the reversal system" anti-injury, promoting repair, anti-mutation, cut lesion"ConclusionsThis study constructed a non-invasive, digital gastric biopsy marking techniques. It depends on the computer image guide and gastric 3D digital model. This technique which is named as computer simulated biopsy marking can be used for the lifetime. Secondary, combined with the patient’s age, COL4A6, COL1A1, CST1, HOXC6 expressed in gastric mucosa are used to establish a discrimination model which can be used to differentiate gastric cancer from gastric precancerous lesions and superficial gastritis. The discriminant accuracy is very high. With the improvement of the bioinformatics digital platform for precancerous lesions, more and more over-expressed genes, RNA and proteins will be involved and be used to discriminate gastric cancer. The accuracy of the model will be higher and higher. The early warning accuracy of precancerous lesions will be significantly improved. Third, the reversion programs of gastric pre-cancerous lesions were compared in this paper and the program with misoprostol and folic acid is superior to other treatment options and is the first choice currently. The individual reversion treatment system for precancerous lesions is also be suggested and established.