BACKGROUND Congenital short bowel syndrome(SBS)associated with malrotation,gut volvulus and jejuno-ileal atresia is a very rare condition.It is a severe challenge for surgeons to preserve residual ischemic bowel segme...BACKGROUND Congenital short bowel syndrome(SBS)associated with malrotation,gut volvulus and jejuno-ileal atresia is a very rare condition.It is a severe challenge for surgeons to preserve residual ischemic bowel segment in the management of short bowel syndrome,especially in neonates.CASE SUMMARY We report a newborn baby with gut malrotation associated with jejuno-ileal atresia,congenital SBS and jejunal volvulus.Hematemesis and abdominal distention were noted.At laparotomy,malrotation associated with jejuno-ileal atresia,congenital SBS and jenunal volvulus was confirmed.The total length of the small bowel was 63 cm with proximal jejunal bowel segment measuring 38 cm,including 18 cm necrotic segment below the Treitz’s ligament and 20 cm severe ischemic segment.The distal part of the small bowel was 25 cm in length and only about 0.8 cm in diameter.Ladd’s procedure,necrotic segment resection and end-to-back duodeno-ileal anastomosis were performed.The residual severe ischemic jejunum was preserved with single proximal stoma and distal end closure.Three months later,to restore the continuity of the isolated gut segment,end-to-end duodeno-jejunal and jejuno-ileal anastomosis was performed.The entire functional small bowel length increased to 80 cm.Intravenous fluid therapy and parenteral nutrition were discontinued on the 10th day postoperatively.Twelve months later,her body weight was 9.5 kg.CONCLUSION Isolation of severe ischemic bowel segment and staged anastomosis to restore the gut continuity for infants with SBS are safe and feasible.展开更多
AIM:To determine whether magnified observation of short-segment Barrett’s esophagus(BE)is useful for the detection of specialized intestinal metaplasia(SIM).METHODS:Thirty patients with suspected short-segment BE und...AIM:To determine whether magnified observation of short-segment Barrett’s esophagus(BE)is useful for the detection of specialized intestinal metaplasia(SIM).METHODS:Thirty patients with suspected short-segment BE underwent magnifying endoscopy up to×80.The magnified images were analyzed with respect to their pit-patterns,which were simultaneously classified into five epithelial types[Ⅰ(small round),Ⅱ(straight),Ⅲ(long oval),Ⅳ(tubular),Ⅴ(villous)]by Endo’s classification.Then,a 0.5%solution of methylene blue(MB)was sprayed over columnar mucosa.The patterns of the magnified image and MB staining were analyzed.Biopsies were obtained from the regions previously observed by magnifying endoscopy and MB chromoendoscopy.RESULTS:Three of five patients with a typeⅤ(villous)epithelial pattern had SIM,whereas 21 patients with a non-typeⅤepithelial patterns did not have SIM.The sensitivity,specificity,accuracy,positive predictive value,and negative predictive value of pit-patterns in detecting SIM were 100%,91.3%,92.3%,60%and100%,respectively(P=0.004).Three of the 12 patients with positive MB staining had SIM,whereas 14patients with negative MB staining did not have SIM.The sensitivity,specificity,accuracy,positive predictive value,and negative predictive value of MB staining in detecting SIM were 100%,60.9%,65.4%,25%and100%,respectively(P=0.085).The specificity and accuracy of pit-pattern evaluation were significantly superior compared with MB staining for detecting SIM by comparison with the exact McNemar’s test(P=0.0391).CONCLUSION:The magnified observation of a shortsegment BE according to the mucosal pattern and its classification can be predictive of SIM.展开更多
The size and shape of the effective test area are crucial to consider when short-crested waves are created by segmented wavemakers. The range of the effective test area of short-crested waves simulated by two-sided se...The size and shape of the effective test area are crucial to consider when short-crested waves are created by segmented wavemakers. The range of the effective test area of short-crested waves simulated by two-sided segmented wavemakers is analyzed in this paper. The experimental investigation on the wave field distribution of short-crested waves generated by two-sided segmented wavemakers is conducted by using an array of wave gauges. Wave spectra and directional spreading function are analyzed and the results show that when the main direction is at a certain angle with the normal line of wave generators, the wave field of 3D short-crested waves generated by two-sided segmented wavemakers has good spatial uniformity within the model test area. The effective test area can provide good wave environments for seakeeping model tests of various ocean engineering structures in the deep ocean engineering basin.展开更多
针对畜禽疫病文本语料匮乏、文本内包含大量疫病名称及短语等未登录词问题,提出了一种结合词典匹配的BERT-BiLSTM-CRF畜禽疫病文本分词模型。以羊疫病为研究对象,构建了常见疫病文本数据集,将其与通用语料PKU结合,利用BERT(Bidirectiona...针对畜禽疫病文本语料匮乏、文本内包含大量疫病名称及短语等未登录词问题,提出了一种结合词典匹配的BERT-BiLSTM-CRF畜禽疫病文本分词模型。以羊疫病为研究对象,构建了常见疫病文本数据集,将其与通用语料PKU结合,利用BERT(Bidirectional encoder representation from transformers)预训练语言模型进行文本向量化表示;通过双向长短时记忆网络(Bidirectional long short-term memory network,BiLSTM)获取上下文语义特征;由条件随机场(Conditional random field,CRF)输出全局最优标签序列。基于此,在CRF层后加入畜禽疫病领域词典进行分词匹配修正,减少在分词过程中出现的疫病名称及短语等造成的歧义切分,进一步提高了分词准确率。实验结果表明,结合词典匹配的BERT-BiLSTM-CRF模型在羊常见疫病文本数据集上的F1值为96.38%,与jieba分词器、BiLSTM-Softmax模型、BiLSTM-CRF模型、未结合词典匹配的本文模型相比,分别提升11.01、10.62、8.3、0.72个百分点,验证了方法的有效性。与单一语料相比,通用语料PKU和羊常见疫病文本数据集结合的混合语料,能够同时对畜禽疫病专业术语及疫病文本中常用词进行准确切分,在通用语料及疫病文本数据集上F1值都达到95%以上,具有较好的模型泛化能力。该方法可用于畜禽疫病文本分词。展开更多
文摘BACKGROUND Congenital short bowel syndrome(SBS)associated with malrotation,gut volvulus and jejuno-ileal atresia is a very rare condition.It is a severe challenge for surgeons to preserve residual ischemic bowel segment in the management of short bowel syndrome,especially in neonates.CASE SUMMARY We report a newborn baby with gut malrotation associated with jejuno-ileal atresia,congenital SBS and jejunal volvulus.Hematemesis and abdominal distention were noted.At laparotomy,malrotation associated with jejuno-ileal atresia,congenital SBS and jenunal volvulus was confirmed.The total length of the small bowel was 63 cm with proximal jejunal bowel segment measuring 38 cm,including 18 cm necrotic segment below the Treitz’s ligament and 20 cm severe ischemic segment.The distal part of the small bowel was 25 cm in length and only about 0.8 cm in diameter.Ladd’s procedure,necrotic segment resection and end-to-back duodeno-ileal anastomosis were performed.The residual severe ischemic jejunum was preserved with single proximal stoma and distal end closure.Three months later,to restore the continuity of the isolated gut segment,end-to-end duodeno-jejunal and jejuno-ileal anastomosis was performed.The entire functional small bowel length increased to 80 cm.Intravenous fluid therapy and parenteral nutrition were discontinued on the 10th day postoperatively.Twelve months later,her body weight was 9.5 kg.CONCLUSION Isolation of severe ischemic bowel segment and staged anastomosis to restore the gut continuity for infants with SBS are safe and feasible.
文摘AIM:To determine whether magnified observation of short-segment Barrett’s esophagus(BE)is useful for the detection of specialized intestinal metaplasia(SIM).METHODS:Thirty patients with suspected short-segment BE underwent magnifying endoscopy up to×80.The magnified images were analyzed with respect to their pit-patterns,which were simultaneously classified into five epithelial types[Ⅰ(small round),Ⅱ(straight),Ⅲ(long oval),Ⅳ(tubular),Ⅴ(villous)]by Endo’s classification.Then,a 0.5%solution of methylene blue(MB)was sprayed over columnar mucosa.The patterns of the magnified image and MB staining were analyzed.Biopsies were obtained from the regions previously observed by magnifying endoscopy and MB chromoendoscopy.RESULTS:Three of five patients with a typeⅤ(villous)epithelial pattern had SIM,whereas 21 patients with a non-typeⅤepithelial patterns did not have SIM.The sensitivity,specificity,accuracy,positive predictive value,and negative predictive value of pit-patterns in detecting SIM were 100%,91.3%,92.3%,60%and100%,respectively(P=0.004).Three of the 12 patients with positive MB staining had SIM,whereas 14patients with negative MB staining did not have SIM.The sensitivity,specificity,accuracy,positive predictive value,and negative predictive value of MB staining in detecting SIM were 100%,60.9%,65.4%,25%and100%,respectively(P=0.085).The specificity and accuracy of pit-pattern evaluation were significantly superior compared with MB staining for detecting SIM by comparison with the exact McNemar’s test(P=0.0391).CONCLUSION:The magnified observation of a shortsegment BE according to the mucosal pattern and its classification can be predictive of SIM.
基金financially supported by the National Natural Science Foundation of China(Grant No.51239007)
文摘The size and shape of the effective test area are crucial to consider when short-crested waves are created by segmented wavemakers. The range of the effective test area of short-crested waves simulated by two-sided segmented wavemakers is analyzed in this paper. The experimental investigation on the wave field distribution of short-crested waves generated by two-sided segmented wavemakers is conducted by using an array of wave gauges. Wave spectra and directional spreading function are analyzed and the results show that when the main direction is at a certain angle with the normal line of wave generators, the wave field of 3D short-crested waves generated by two-sided segmented wavemakers has good spatial uniformity within the model test area. The effective test area can provide good wave environments for seakeeping model tests of various ocean engineering structures in the deep ocean engineering basin.
文摘针对畜禽疫病文本语料匮乏、文本内包含大量疫病名称及短语等未登录词问题,提出了一种结合词典匹配的BERT-BiLSTM-CRF畜禽疫病文本分词模型。以羊疫病为研究对象,构建了常见疫病文本数据集,将其与通用语料PKU结合,利用BERT(Bidirectional encoder representation from transformers)预训练语言模型进行文本向量化表示;通过双向长短时记忆网络(Bidirectional long short-term memory network,BiLSTM)获取上下文语义特征;由条件随机场(Conditional random field,CRF)输出全局最优标签序列。基于此,在CRF层后加入畜禽疫病领域词典进行分词匹配修正,减少在分词过程中出现的疫病名称及短语等造成的歧义切分,进一步提高了分词准确率。实验结果表明,结合词典匹配的BERT-BiLSTM-CRF模型在羊常见疫病文本数据集上的F1值为96.38%,与jieba分词器、BiLSTM-Softmax模型、BiLSTM-CRF模型、未结合词典匹配的本文模型相比,分别提升11.01、10.62、8.3、0.72个百分点,验证了方法的有效性。与单一语料相比,通用语料PKU和羊常见疫病文本数据集结合的混合语料,能够同时对畜禽疫病专业术语及疫病文本中常用词进行准确切分,在通用语料及疫病文本数据集上F1值都达到95%以上,具有较好的模型泛化能力。该方法可用于畜禽疫病文本分词。
