Deep learning models have been shown to have great advantages in answer selection tasks.The existing models,which employ encoder-decoder recurrent neural network(RNN),have been demonstrated to be effective.However,the...Deep learning models have been shown to have great advantages in answer selection tasks.The existing models,which employ encoder-decoder recurrent neural network(RNN),have been demonstrated to be effective.However,the traditional RNN-based models still suffer from limitations such as 1)high-dimensional data representation in natural language processing and 2)biased attentive weights for subsequent words in traditional time series models.In this study,a new answer selection model is proposed based on the Bidirectional Long Short-Term Memory(Bi-LSTM)and attention mechanism.The proposed model is able to generate the more effective question-answer pair representation.Experiments on a question answering dataset that includes information from multiple fields show the great advantages of our proposed model.Specifically,we achieve a maximum improvement of 3.8%over the classical LSTM model in terms of mean average precision.展开更多
Background:Type 1 autoimmune pancreatitis(AIP)is the pancreatic manifestation of IgG4-related disease.However,this benign disease can result in the peripancreatic vascular involvement(PVI)on occasion,which increases t...Background:Type 1 autoimmune pancreatitis(AIP)is the pancreatic manifestation of IgG4-related disease.However,this benign disease can result in the peripancreatic vascular involvement(PVI)on occasion,which increases the difficulty of diagnosis and treatment of this clinical entity as well as for differentiating it from pancreatic malignancies.Methods:We retrospectively reviewed the information on demographics,clinical presentation,laboratory,imaging and endoscopic findings of 101 hospitalized patients with type 1 AIP treated in our department.All the patients were divided into non-PVI and PVI groups according to the first hospitalized medical data.Univariate and multivariate analyses were performed to analyse the potential predictive parameter(s)of PVI in AIP patients.Results:Among the 101 type 1 AIP patients,52(51.5%)exhibited PVI,with a male/female ratio 5.5:1.Their average age was 58.37±8.68 years old.Univariate analysis revealed that the location of pancreatitis lesions,including the pancreatic tail(P=0.010),the presence of splenomegaly(P=0.001)and the white blood cell(WBC)number in peripheral blood(P=0.020),were significantly associated with PVI.The location of pancreatitis lesions,including the pancreatic tail(P=0.023),and the presence of splenomegaly(P=0.010)were found to be independent predictors of the development of PVI by a multivariable regression analysis.A total of 18 out of 25 patients in PVI group who underwent corticosteroid treatment and no less than 6 months radiological follow-up showed improvement in vascular lesions,and no case exhibited exacerbation of PVI lesions during follow-up.Of 36 patients in non-PVI group who were followed up for no less than 6 months,only one case exhibited PVI.Conclusions:This retrospective study demonstrated that type 1 AIP was associated with a high proportion of PVI.Pancreatic tail involvement and splenomegaly may predict the PVI in type 1 AIP.PVI lesions are reversible in a subset of patients.展开更多
基金This work was supported in part by the National Natural Science Foundation of China under Grant 61572326,and Grant 61802258the Natural Science Foundation of Shanghai under Grant 18ZR1428300the Shanghai Committee of Science and Technology under Grant 17070502800 and Grant 16JC1403000.
文摘Deep learning models have been shown to have great advantages in answer selection tasks.The existing models,which employ encoder-decoder recurrent neural network(RNN),have been demonstrated to be effective.However,the traditional RNN-based models still suffer from limitations such as 1)high-dimensional data representation in natural language processing and 2)biased attentive weights for subsequent words in traditional time series models.In this study,a new answer selection model is proposed based on the Bidirectional Long Short-Term Memory(Bi-LSTM)and attention mechanism.The proposed model is able to generate the more effective question-answer pair representation.Experiments on a question answering dataset that includes information from multiple fields show the great advantages of our proposed model.Specifically,we achieve a maximum improvement of 3.8%over the classical LSTM model in terms of mean average precision.
基金National Natural Science Foundation of China,Joint Fund Project(grant No.U20A600)National Key R&D Program of China(grant No.2020YFC2002702)+1 种基金National Key Technology Research and Development Program(grant No.2016YFC1302802)Beijing Municipal Science and Technology Project(grant No.Z181100001618013).
文摘Background:Type 1 autoimmune pancreatitis(AIP)is the pancreatic manifestation of IgG4-related disease.However,this benign disease can result in the peripancreatic vascular involvement(PVI)on occasion,which increases the difficulty of diagnosis and treatment of this clinical entity as well as for differentiating it from pancreatic malignancies.Methods:We retrospectively reviewed the information on demographics,clinical presentation,laboratory,imaging and endoscopic findings of 101 hospitalized patients with type 1 AIP treated in our department.All the patients were divided into non-PVI and PVI groups according to the first hospitalized medical data.Univariate and multivariate analyses were performed to analyse the potential predictive parameter(s)of PVI in AIP patients.Results:Among the 101 type 1 AIP patients,52(51.5%)exhibited PVI,with a male/female ratio 5.5:1.Their average age was 58.37±8.68 years old.Univariate analysis revealed that the location of pancreatitis lesions,including the pancreatic tail(P=0.010),the presence of splenomegaly(P=0.001)and the white blood cell(WBC)number in peripheral blood(P=0.020),were significantly associated with PVI.The location of pancreatitis lesions,including the pancreatic tail(P=0.023),and the presence of splenomegaly(P=0.010)were found to be independent predictors of the development of PVI by a multivariable regression analysis.A total of 18 out of 25 patients in PVI group who underwent corticosteroid treatment and no less than 6 months radiological follow-up showed improvement in vascular lesions,and no case exhibited exacerbation of PVI lesions during follow-up.Of 36 patients in non-PVI group who were followed up for no less than 6 months,only one case exhibited PVI.Conclusions:This retrospective study demonstrated that type 1 AIP was associated with a high proportion of PVI.Pancreatic tail involvement and splenomegaly may predict the PVI in type 1 AIP.PVI lesions are reversible in a subset of patients.