Isogeometric analysis (IGA) is known to showadvanced features compared to traditional finite element approaches.Using IGA one may accurately obtain the geometrically nonlinear bending behavior of plates with functiona...Isogeometric analysis (IGA) is known to showadvanced features compared to traditional finite element approaches.Using IGA one may accurately obtain the geometrically nonlinear bending behavior of plates with functionalgrading (FG). However, the procedure is usually complex and often is time-consuming. We thus put forward adeep learning method to model the geometrically nonlinear bending behavior of FG plates, bypassing the complexIGA simulation process. A long bidirectional short-term memory (BLSTM) recurrent neural network is trainedusing the load and gradient index as inputs and the displacement responses as outputs. The nonlinear relationshipbetween the outputs and the inputs is constructed usingmachine learning so that the displacements can be directlyestimated by the deep learning network. To provide enough training data, we use S-FSDT Von-Karman IGA andobtain the displacement responses for different loads and gradient indexes. Results show that the recognition erroris low, and demonstrate the feasibility of deep learning technique as a fast and accurate alternative to IGA formodeling the geometrically nonlinear bending behavior of FG plates.展开更多
BACKGROUND Gastric cancer has a high incidence and fatality rate,and surgery is the preferred course of treatment.Nonetheless,patient survival rates are still low,and the incidence of major postoperative complications...BACKGROUND Gastric cancer has a high incidence and fatality rate,and surgery is the preferred course of treatment.Nonetheless,patient survival rates are still low,and the incidence of major postoperative complications cannot be disregarded.The systemic inflammatory response,nutritional level,and coagulation status are key factors affecting the postoperative recovery and prognosis of gastric cancer patients.The systemic inflammatory response index(SIRI)and the albumin fibrinogen ratio(AFR)are two valuable comprehensive indicators of the severity and prognosis of systemic inflammation in various medical conditions.AIM To assess the clinical importance and prognostic significance of the SIRI scores and the AFR on early postoperative outcomes in patients undergoing radical gastric cancer surgery.METHODS We conducted a retrospective analysis of the clinicopathological characteristics and relevant laboratory indices of 568 gastric cancer patients from January 2018 to December 2019.We calculated and compared two indicators of inflammation and then examined the diagnostic ability of combined SIRI and AFR values for serious early postoperative complications.We scored the patients and categorized them into three groups based on their SIRI and AFR levels.COX analysis was used to compare the three groups of patients the prognostic value of various preoperative SIRI-AFR scores for 5-year overall survival(OS)and disease-free survival(DFS).RESULTS SIRI-AFR scores were an independent risk factor for prognosis[OS:P=0.004;hazards ratio(HR)=3.134;DFS:P<0.001;HR=3.543]and had the highest diagnostic power(area under the curve:0.779;95%confidence interval:0.737-0.820)for early serious complications in patients with gastric cancer.The tumor-node-metastasis stage(P=0.001),perioperative transfusion(P=0.044),positive carcinoembryonic antigen(P=0.014)findings,and major postoperative complications(P=0.011)were factors associated with prognosis.CONCLUSION Preoperative SIRI and AFR values were significantly associated with early postoperative survival and the occurrence of severe complications in gastric cancer patients.展开更多
基金the National Natural Science Foundation of China(NSFC)under Grant Nos.12272124 and 11972146.
文摘Isogeometric analysis (IGA) is known to showadvanced features compared to traditional finite element approaches.Using IGA one may accurately obtain the geometrically nonlinear bending behavior of plates with functionalgrading (FG). However, the procedure is usually complex and often is time-consuming. We thus put forward adeep learning method to model the geometrically nonlinear bending behavior of FG plates, bypassing the complexIGA simulation process. A long bidirectional short-term memory (BLSTM) recurrent neural network is trainedusing the load and gradient index as inputs and the displacement responses as outputs. The nonlinear relationshipbetween the outputs and the inputs is constructed usingmachine learning so that the displacements can be directlyestimated by the deep learning network. To provide enough training data, we use S-FSDT Von-Karman IGA andobtain the displacement responses for different loads and gradient indexes. Results show that the recognition erroris low, and demonstrate the feasibility of deep learning technique as a fast and accurate alternative to IGA formodeling the geometrically nonlinear bending behavior of FG plates.
基金the National Natural Science Foundation of China,No.8236110677Central to guide local scientific and Technological Development,No.ZYYDDFFZZJ-1+1 种基金Natural Science Foundation of Gansu Province,China,No.18JR2RA033Gansu Da Vinci Robot High-End Diagnosis and Treatment Team Construction Project,National Key Research and Development Program,No.2020RCXM076.
文摘BACKGROUND Gastric cancer has a high incidence and fatality rate,and surgery is the preferred course of treatment.Nonetheless,patient survival rates are still low,and the incidence of major postoperative complications cannot be disregarded.The systemic inflammatory response,nutritional level,and coagulation status are key factors affecting the postoperative recovery and prognosis of gastric cancer patients.The systemic inflammatory response index(SIRI)and the albumin fibrinogen ratio(AFR)are two valuable comprehensive indicators of the severity and prognosis of systemic inflammation in various medical conditions.AIM To assess the clinical importance and prognostic significance of the SIRI scores and the AFR on early postoperative outcomes in patients undergoing radical gastric cancer surgery.METHODS We conducted a retrospective analysis of the clinicopathological characteristics and relevant laboratory indices of 568 gastric cancer patients from January 2018 to December 2019.We calculated and compared two indicators of inflammation and then examined the diagnostic ability of combined SIRI and AFR values for serious early postoperative complications.We scored the patients and categorized them into three groups based on their SIRI and AFR levels.COX analysis was used to compare the three groups of patients the prognostic value of various preoperative SIRI-AFR scores for 5-year overall survival(OS)and disease-free survival(DFS).RESULTS SIRI-AFR scores were an independent risk factor for prognosis[OS:P=0.004;hazards ratio(HR)=3.134;DFS:P<0.001;HR=3.543]and had the highest diagnostic power(area under the curve:0.779;95%confidence interval:0.737-0.820)for early serious complications in patients with gastric cancer.The tumor-node-metastasis stage(P=0.001),perioperative transfusion(P=0.044),positive carcinoembryonic antigen(P=0.014)findings,and major postoperative complications(P=0.011)were factors associated with prognosis.CONCLUSION Preoperative SIRI and AFR values were significantly associated with early postoperative survival and the occurrence of severe complications in gastric cancer patients.
