The Prague C and M Criteria have been developed for the objective endoscopic diagnosis of Barrett's esophagus(BE).BE arises between the squamocolumnar junction and the gastroesophageal junction at the proximal mar...The Prague C and M Criteria have been developed for the objective endoscopic diagnosis of Barrett's esophagus(BE).BE arises between the squamocolumnar junction and the gastroesophageal junction at the proximal margin of the gastric folds.In this study,we reported that 43.0% of the subjects examined were diagnosed with BE based on the Prague C and M Criteria.Previous criticism by John Dent proposed that our data should be considered invalid because the prevalence of BE reported in our study was extraordinarily high and discordant with previous studies.Dent predicted that the position of the gastroesophageal junction in our study was judged to be lower than the actual position due to the effacement of the proximal ends of the gastric folds because of the routine use of a high degree of air distension during typical Japanese endoscopic examinations.The endoscopic evaluation of the superior gastric folds is certainly influenced by the degree of air distension of the esophagus.However,in our study,the proximal limit of the gastric mucosal folds was prospectively imaged while the oesophagus was minimally insufflated.Then,under a high level of air distension,the distal ends of the palisade-shaped longitudinal vessels were imaged because they are more easily observed when distended.In the majority of patients,the distal ends of the palisade-shaped longitudinal vessels correspond to the proximal limit of the gastric mucosal folds.Our endoscopic evaluation was appropriately performed according to the Prague C and M Criteria.We suspect that the high prevalence of BE in our study may be due to the inclusion of ultrashort-segment BE,which defines BE with an affected mucosal length under 5 mm,in our positive results.展开更多
Input selection is probably one of the most critical decision issues in neural network designing, because it has a great impact on forecasting performance. Among the many applications of artificial neural networks to ...Input selection is probably one of the most critical decision issues in neural network designing, because it has a great impact on forecasting performance. Among the many applications of artificial neural networks to finance, time series forecasting is perhaps one of the most challenging issues. Considering the features of neural networks, we propose a general approach called Autocorrelation Criterion (AC) to determine the inputs variables for a neural network. The purpose is to seek optimal lag periods, which are more predictive and less correlated. AC is a data-driven approach in that there is no prior assumptiona bout the models for time series under study. So it has extensive applications and avoids a lengthy experimentation and tinkering in input selection. We apply the approach to the determination of input variables for foreign exchange rate forecasting and conductcomparisons between AC and information-based in-sample model selection criterion. The experiment results show that AC outperforms information-based in-sample model selection criterion.展开更多
文摘The Prague C and M Criteria have been developed for the objective endoscopic diagnosis of Barrett's esophagus(BE).BE arises between the squamocolumnar junction and the gastroesophageal junction at the proximal margin of the gastric folds.In this study,we reported that 43.0% of the subjects examined were diagnosed with BE based on the Prague C and M Criteria.Previous criticism by John Dent proposed that our data should be considered invalid because the prevalence of BE reported in our study was extraordinarily high and discordant with previous studies.Dent predicted that the position of the gastroesophageal junction in our study was judged to be lower than the actual position due to the effacement of the proximal ends of the gastric folds because of the routine use of a high degree of air distension during typical Japanese endoscopic examinations.The endoscopic evaluation of the superior gastric folds is certainly influenced by the degree of air distension of the esophagus.However,in our study,the proximal limit of the gastric mucosal folds was prospectively imaged while the oesophagus was minimally insufflated.Then,under a high level of air distension,the distal ends of the palisade-shaped longitudinal vessels were imaged because they are more easily observed when distended.In the majority of patients,the distal ends of the palisade-shaped longitudinal vessels correspond to the proximal limit of the gastric mucosal folds.Our endoscopic evaluation was appropriately performed according to the Prague C and M Criteria.We suspect that the high prevalence of BE in our study may be due to the inclusion of ultrashort-segment BE,which defines BE with an affected mucosal length under 5 mm,in our positive results.
基金This research is partially supported by Chinese Academy of SciencesNational Science Foundation of ChinaJapan Society for the Promotion of Science.
文摘Input selection is probably one of the most critical decision issues in neural network designing, because it has a great impact on forecasting performance. Among the many applications of artificial neural networks to finance, time series forecasting is perhaps one of the most challenging issues. Considering the features of neural networks, we propose a general approach called Autocorrelation Criterion (AC) to determine the inputs variables for a neural network. The purpose is to seek optimal lag periods, which are more predictive and less correlated. AC is a data-driven approach in that there is no prior assumptiona bout the models for time series under study. So it has extensive applications and avoids a lengthy experimentation and tinkering in input selection. We apply the approach to the determination of input variables for foreign exchange rate forecasting and conductcomparisons between AC and information-based in-sample model selection criterion. The experiment results show that AC outperforms information-based in-sample model selection criterion.
