A neural network model with a classical annotation method has been used on the EXL-50tokamak to predict impending disruption.However,the results revealed issues of overfitting and overconfidence in predictions caused ...A neural network model with a classical annotation method has been used on the EXL-50tokamak to predict impending disruption.However,the results revealed issues of overfitting and overconfidence in predictions caused by inaccurate labeling.To mitigate these issues,an improved training framework has been proposed.In this approach,soft labels from previous training serve as teachers to supervise the further learning process;this has lead to a significant improvement in predictive model performance.Notably,this enhancement is primarily attributed to the coupling effect of the soft labels and correction mechanism.This improved training framework introduces an instance-specific label smoothing method,which reflects a more nuanced model assessment on the likelihood of a disruption.It presents a possible solution to effectively address the challenges associated with accurate labeling across different machines.展开更多
No direct comparison has been performed between different programmed cell death-1(PD-1)inhibitors for first-line treatment in patients with advanced non-small cell lung cancer(NSCLC).The feasibility of using PD-L1-exp...No direct comparison has been performed between different programmed cell death-1(PD-1)inhibitors for first-line treatment in patients with advanced non-small cell lung cancer(NSCLC).The feasibility of using PD-L1-expression-guided immunotherapy remains unknown.In this open-label,phase 2 study(NCT04252365),patients with advanced NSCLC without EGFR or ALK alterations were randomized(1:1)to receive sintilimab or pembrolizumab monotherapy(PD-L1 expression≥50%),or sintilimab or pembrolizumab plus platinum-based chemotherapy(PD-L1 expression<50%).The sample size was calculated by optimal two-stage design.The primary endpoint was the objective response rate(ORR).The study included 71 patients(sintilimab arms,n=35;pembrolizumab arms,n=36)and met its primary endpoint,with a confirmed ORR of 51.4%(18/35)in the sintilimab arms.The confirmed ORR(95%confidence interval)was 46.2%(19.2%,74.9%)and 42.9%(17.7%,71.1%)for patients treated with sintilimab and pembrolizumab monotherapy;and 54.5%(32.2%,75.6%)and 45.4%(24.4%,67.8%)for those treated with sintilimab-and pembrolizumab-based combination therapies.The median progression-free survival was6.9 versus 8.1 months for all sintilimab-treated versus all pembrolizumab-treated patients,respectively,in which it was 7.6 versus 11.0 months in monotherapy and 7.4 versus 7.1 months in combination therapies.The median overall survival was 14.9 versus 21.3 months for all sintilimab-treated versus all pembrolizumab-treated patients,respectively,in which it was 14.9 versus 22.6 months in monotherapy and 14.7 versus 17.3 months in combination therapies.Treatment-related adverse events were consistent with safety outcomes of monotherapy and combination therapy in previous phase III studies.However,the incidence of rash was higher with sintilimab than pembrolizumab monotherapy.This is the first prospective phase 2 study to directly compare two anti-PD-1 antibodies as first-line treatment in advanced NSCLC.Sintilimab was efficacious and well-tolerated irrespective of PD-L1 expression level in 展开更多
Multiphoton microscopy is the enabling tool for biomedical research,but the aberrations of biological tissues have limited its imaging performance.Adaptive optics(AO)has been developed to partially overcome aberration...Multiphoton microscopy is the enabling tool for biomedical research,but the aberrations of biological tissues have limited its imaging performance.Adaptive optics(AO)has been developed to partially overcome aberration to restore imaging performance.For indirect AO,algorithm is the key to its successful implementation.Here,based on the fact that indirect AO has an analogy to the black-box optimization problem,we successfully apply the covariance matrix adaptation evolution strategy(CMA-ES)used in the latter,to indirect AO in multiphoton microscopy(MPM).Compared with the traditional genetic algorithm(GA),our algorithm has a greater improvement in convergence speed and convergence accuracy,which provides the possibility of realizing real-time dynamic aberration correction for deep in vivo biological tissues.展开更多
基金supported by National Natural Science Foundation of China(Nos.12175277 and 11975271)the National Key R&D Program of China(No.2022YFE 03050003)。
文摘A neural network model with a classical annotation method has been used on the EXL-50tokamak to predict impending disruption.However,the results revealed issues of overfitting and overconfidence in predictions caused by inaccurate labeling.To mitigate these issues,an improved training framework has been proposed.In this approach,soft labels from previous training serve as teachers to supervise the further learning process;this has lead to a significant improvement in predictive model performance.Notably,this enhancement is primarily attributed to the coupling effect of the soft labels and correction mechanism.This improved training framework introduces an instance-specific label smoothing method,which reflects a more nuanced model assessment on the likelihood of a disruption.It presents a possible solution to effectively address the challenges associated with accurate labeling across different machines.
基金supported by the Guangdong Provincial Key Lab of Translational Medicine in Lung Cancer(2017B030314120)the Guangdong Provincial People’s Hospital Scientific Research Funds for Leading Medical Talents in Guangdong Province(KJ012019426)+2 种基金the National Natural Science Foundation of China(82072562 and 82202997)the China Postdoctoral Science Foundation(2021M701422)the High-Level Hospital Construction Project(DFJH201810).
文摘No direct comparison has been performed between different programmed cell death-1(PD-1)inhibitors for first-line treatment in patients with advanced non-small cell lung cancer(NSCLC).The feasibility of using PD-L1-expression-guided immunotherapy remains unknown.In this open-label,phase 2 study(NCT04252365),patients with advanced NSCLC without EGFR or ALK alterations were randomized(1:1)to receive sintilimab or pembrolizumab monotherapy(PD-L1 expression≥50%),or sintilimab or pembrolizumab plus platinum-based chemotherapy(PD-L1 expression<50%).The sample size was calculated by optimal two-stage design.The primary endpoint was the objective response rate(ORR).The study included 71 patients(sintilimab arms,n=35;pembrolizumab arms,n=36)and met its primary endpoint,with a confirmed ORR of 51.4%(18/35)in the sintilimab arms.The confirmed ORR(95%confidence interval)was 46.2%(19.2%,74.9%)and 42.9%(17.7%,71.1%)for patients treated with sintilimab and pembrolizumab monotherapy;and 54.5%(32.2%,75.6%)and 45.4%(24.4%,67.8%)for those treated with sintilimab-and pembrolizumab-based combination therapies.The median progression-free survival was6.9 versus 8.1 months for all sintilimab-treated versus all pembrolizumab-treated patients,respectively,in which it was 7.6 versus 11.0 months in monotherapy and 7.4 versus 7.1 months in combination therapies.The median overall survival was 14.9 versus 21.3 months for all sintilimab-treated versus all pembrolizumab-treated patients,respectively,in which it was 14.9 versus 22.6 months in monotherapy and 14.7 versus 17.3 months in combination therapies.Treatment-related adverse events were consistent with safety outcomes of monotherapy and combination therapy in previous phase III studies.However,the incidence of rash was higher with sintilimab than pembrolizumab monotherapy.This is the first prospective phase 2 study to directly compare two anti-PD-1 antibodies as first-line treatment in advanced NSCLC.Sintilimab was efficacious and well-tolerated irrespective of PD-L1 expression level in
基金supported by the National Natural Science Foundation of China(Nos.62075135 and 61975126)the Science,Technology and Innovation Commission of Shenzhen Municipality(Nos.JCYJ20190808174819083 and JCYJ20190808175201640)。
文摘Multiphoton microscopy is the enabling tool for biomedical research,but the aberrations of biological tissues have limited its imaging performance.Adaptive optics(AO)has been developed to partially overcome aberration to restore imaging performance.For indirect AO,algorithm is the key to its successful implementation.Here,based on the fact that indirect AO has an analogy to the black-box optimization problem,we successfully apply the covariance matrix adaptation evolution strategy(CMA-ES)used in the latter,to indirect AO in multiphoton microscopy(MPM).Compared with the traditional genetic algorithm(GA),our algorithm has a greater improvement in convergence speed and convergence accuracy,which provides the possibility of realizing real-time dynamic aberration correction for deep in vivo biological tissues.
