| Colorectal cancer(CRC)is one of the most common malignant tumors in the world and the third largest cancer in China.The key to the prevention and treatment of CRC is early detection and early treatment,but because early CRC patients have no specific clinical manifestations,early diagnosis is difficult,which seriously affects patients' quality of life and 5-year survival rate.In Europe and the United States,due to the implementation of colorectal cancer screening programs in the population,the morbidity and mortality of colorectal cancer continue to decrease.At present,the early diagnosis rate of CRC in China is significantly lower than in Europe and the United States.Colonoscopy can not only detect and observe the lesion directly,but also can directly intervene under the colonoscope to prevent the formation and development of CRC,it is considered as the gold standard for CRC screening.However,there is still a missed diagnosis of the lesion that cannot be ignored.Therefore,it is of great practical significance to further improve the efficiency of colonoscopy and reduce the adenoma miss rate during the examination.This subject starts from improving the quality of bowel preparation,explores interventions to optimize the bowel preparation of colonoscopy,and addresses the strategic issues of colorectal polyps treatment during colonoscopy,and jointly develops artificial intelligence(AI)assistive technology with computer software,and it is used in clinic to assist endoscopists to identify the pathological nature of colorectal polyps..Section ?: The effect of a Low-residue Food for Special Medical Purpose on the quality of bowel preparation for colonoscopy: a single-blind,randomized controlled,multi-center clinical trialBackground and objective: Adequate bowel preparation is the basis and prerequisite for high-quality colonoscopy.Diet restriction is an important step in bowel preparation.Previous research has shown that compared with a self-prepared low-residue diet,a pre-packaged low-residue diet can significantly improve patient compliance and tolerance and help improve the quality of bowel preparation.However,there is currently no widely accepted pre-packaged diet for bowel preparation for colonoscopy,and the pre-packaged diet used in previous studies is traditional unrefined food.The purpose of this study was to investigate the value of a low-residue food for special medical purpose(FSMP)for bowel preparation for colonoscopy.Methods: This is a single-blind,randomized controlled study of a multicenter prospective endoscopic physician.The subjects were randomly divided into the experimental group or the control group.Patients it the experimental group took only low-residue special medical formula food one day before the colonoscopy.Patients it the control group should prepare low-residue diet for themself.The two groups adopted the same bowel cleansing scheme.The outcomes included adequate bowel preparation rate,bowel preparation score,adenoma detection rate(ADR),polyp detection rate(PDR),diet restriction compliance rate,bowel preparation completion rate,cecal intubation rate and time,incidence of adverse reactions,and tolerance.Results: A total of 183 subjects were enrolled,93 in the test group and 90 in the control group.There were no significant differences in baseline characteristics between the two groups.The results of outcomes are as follows: adequate bowel preparation rate(97.8% Vs.91.1%,P = 0.093);bowel preparation score(8.05 ± 0.89 Vs.7.37 ± 1.34,P < 0.001);ADR(15.1% Vs.18.9%,P = 0.489);PDR(36.6% Vs.37.8%,P = 0.865);diet restriction compliance rate(82.8 % Vs.63.3%,P = 0.003);bowel preparation completion rate,(96.8% Vs.98.9%,P = 0.328);cecal intubation rate(both 100%)and time of insertion(5.43 ± 4.45 Vs 5.79 ± 4.55,P = 0.587);preoperative adverse reaction rate(3.2% Vs.11.1%,P = 0.038);all three outcome of tolerance: P <0.001.Conclusion: The low-residue special medical formula food shows higher quality of bowel preparation,higher diet restriction compliance rate,lower incidence of adverse reactions,better patient tolerance,and can be used as standard pre-packaging diet for colonoscopy.Section ?: Effectiveness and safety of oral sodium sulfate potassium magnesium solution for bowel preparation for colonoscopy: a randomized controlled,double-blind,positive parallel controlled,multicenter clinical studyBackground and objective: Polyethylene glycol(PEG)is currently the most widely used for bowel preparation,but the liquid volume of oral PEG is and it has a bad taste,so some patients fail to take enough laxatives to complete the bowel preparation.Sodium potassium magnesium is an improved preparation based on magnesium sulfate,which can reduce liquid intake,taste good,and be cheaper.The purpose of this study was to investigate the effect,safety and tolerance of potassium sodium magnesium sulfate oral solution in bowel preparation for colonoscopy.Methods: This trial is a randomized,double-blind,positive,parallel randomized,multi-center,non-inferiority clinical trial.In the experimental group,bowel preparation was performed using sodium potassium sulfate and magnesium,and in the control group,PEG was use.The two groups were compared in terms of adequate bowel preparation rate,bowel preparation score,rug acceptance,medication compliance,incidence of adverse events rate,and tolerance.Results: A total of 348 subjects were enrolled,174 in the test group and 174 in the control group.There were no significant differences in baseline characteristics between the two groups.The results of each outcomes are as follows: the adequate bowel preparation rate in the experimental group is not inferior to that in the control group(A%-B% < 0.1);the total score of BBPS,drug acceptance,drug compliance,colonoscopy completion,and adverse events rate were not Significant differences.Conclusions: The oral sodium sulfate potassium magnesium solution showed no inferior bowel preparation quality than PEG,and the subjects were well tolerated,safe,and cheaper,and it could be used as an effective bowel preparation laxative before colonoscopy.Section ?: Establishment and preliminary clinical verification of an artificial intelligence classification system of colonic polypsBackground and objective: There have been many reports on the study of the combination of artificial intelligence(AI)and colonoscopy,but there have been no reports of validation in a clinical colonoscopy.The purpose of this study is to use clinical big data to train and establish an AI-assisted polyp classification system,and apply it to a clinical colonoscopy real-time environment for verification.Methods: Polyp images(white light and NBI)of each participating research center were prospectively collected and labelled for training and testing the AI system.The established AI system was installed in the endoscopic center,combined with clinical colonoscopy.with pathological results as the standard,and adenomatous polyps as positive results,the system's performance of classification colorectal polyps was initially verified.Results: A total of 15,441 qualified images of 3,946 polyps were collected,of which 9,109 were white light images and 6,332 were NBI images.At the laboratory level,the performance of the system is excellent: the sensitivity and specificity of white light and NPI are more than 90%.The preliminary clinical verification included a total of 56 colonoscopy patients with a total of 78 polyps.The sensitivity,specificity,positive predictive value,negative predictive value,and accuracy of NBI model AI system were 78.4%,87.2%,85.3%,81.0 %,82.9%;and 70.3%,82.1%,78.8%,74.4%,76.3% in white light mode,respectively,and there is no significant difference with the pathological results and the results of endoscopic physicians,but the level of consistency with pathological results is fair,with kappa values 0.581(NBI)and 0.481(white light).Research conclusions: An AI-assisted classification system for colorectal polyps that can be used in clinical practice is established to help endoscopists,especially beginners,make decisions when dealing with colorectal polyps,but its accuracy needs to be further improved. |
PDF Full Text Request