背景与目的:弥漫大B细胞淋巴瘤(diffuse large B-cell lymphoma,DLBCL)的生发中心B细胞样(germinal center B-cell-like,GCB)亚型和非GCB(non-GCB)亚型在患者预后和治疗上存在差异,但目前依赖有创病理学检查。本研究基于多参数MRI构建...背景与目的:弥漫大B细胞淋巴瘤(diffuse large B-cell lymphoma,DLBCL)的生发中心B细胞样(germinal center B-cell-like,GCB)亚型和非GCB(non-GCB)亚型在患者预后和治疗上存在差异,但目前依赖有创病理学检查。本研究基于多参数MRI构建影像组学和深度学习模型,旨在于术前无创性区分这两种亚型。方法:本研究回顾性分析2013年3月—2024年12月在复旦大学附属华山医院及外院经病理学检查确诊的DLBCL患者。使用多参数MRI扫描数据,结合4种影像组学机器学习[支持向量机(support vector machine,SVM)、逻辑回归(logistic regression,LR)、高斯过程(Gaussian process,GP)和朴素贝叶斯(Naive Bayes,NB)]和3种深度学习[密集连接卷积网络121(densely-connected convolutional networks 121,DenseNet121)、残差网络101(residual network 101,ResNet101)和高效网络B5(Efficient Net-b5)]建立DLBCL亚型分类模型。此外,两名经验不同的放射科医师在盲法下基于MRI图像独立分类DLBCL。模型和医师的诊断性能均通过接收者操作特征曲线下面积(area under the curve,AUC)、准确度(accuracy,ACC)和F1分数(F1-score,F1)等指标进行量化评估,以衡量其区分GCB和non-GCB亚型的能力。本研究经复旦大学附属华山医院伦理委员会批准(KY2024-663),所有患者均知情同意。结果:本研究共纳入173例患者(GCB型55例,non-GCB型118例)。影像组学和深度学习方法能有效地区分DLBCL亚型。其中,GP影像组学模型(基于T1-CE+T2-FLAIR+ADC序列)和DenseNet121深度学习模型(基于T1-CE+T2-FLAIR+ADC序列)表现最佳,在内部验证集上分别取得优异性能(GP:AUC=0.900,ACC=0.896,F1=0.840;DenseNet121:AUC=0.846,ACC=0.854,F1=0.774),并在外部验证集上保持稳健。并且,最优AI模型的分类效能优于经验丰富的放射科医师(医师最高AUC=0.678)。结论:基于多参数MRI特征的影像组学与深度学习模型可有效地鉴别DLBCL的GCB与non-GCB亚型。其中,GP与DenseNet121模型在处理复杂图像数据、特别是融合多序列特征组进行亚型分类时,呈现出优异的性能。展开更多
This study proposes an alternative calculation mode for stresses on the slip surface(SS).The calculation of the normal stress(NS)on the SS involves examining its composition and expanding its unknown using the Taylor ...This study proposes an alternative calculation mode for stresses on the slip surface(SS).The calculation of the normal stress(NS)on the SS involves examining its composition and expanding its unknown using the Taylor series.This expansion enables the reasonable construction of a function describing the NS on the SS.Additionally,by directly incorporating the nonlinear Generalized Hoke-Brown(GHB)strength criterion and utilizing the slope factor of safety(FOS)definition,a function of the shear stress on the SS is derived.This function considers the mutual feedback mechanism between the NS and strength parameters of the SS.The stress constraints conditions are then introduced at both ends of the SS based on the spatial stress relation of one point.Determining the slope FOS and stress solution for the SS involves considering the mechanical equilibrium conditions and the stress constraint conditions satisfied by the sliding body.The proposed approach successfully simulates the tension-shear stress zone near the slope top and provides an intuitive description of the concentration effect of compression-shear stress of the SS near the slope toe.Furthermore,compared to other methods,the present method demonstrates superior processing capabilities for the embedded nonlinear GHB strength criterion.展开更多
For rechargeable aqueous zinc-ion batteries(ZIBs),the design of nanocomposites comprised of electrochemically active materials and carbon materials with novel structures has great prom-ise in addressing the issue of e...For rechargeable aqueous zinc-ion batteries(ZIBs),the design of nanocomposites comprised of electrochemically active materials and carbon materials with novel structures has great prom-ise in addressing the issue of electrical conductivity and structural stability in the electrode materials during electrochemical cycling.We report the production of a novel flexible electrode material,by anchoring MnO_(2) nanosheets on a B,N co-doped carbon nanotube ar-ray(BNCNTs)grown on carbon cloth(BNCNTs@MnO_(2)),which was fabricated by in-situ pyrolysis and hydrothermal growth.The generated BNCNTs were strongly bonded to the surface of the car-bon fibers in the carbon cloth which provides both excellent elec-tron transport and ion diffusion,and improves the stability and dur-ability of the cathode.Importantly,the BNCNTs offer more active sites for the hydrothermal growth of MnO_(2),ensuring a uniform dis-tribution.Electrochemical tests show that BNCNTs@MnO_(2) delivers a high specific capacity of 310.7 mAh g^(−1) at 0.1 A g^(−1),along with excellent rate capability and outstanding cycling stability,with a 79.7% capacity retention after 8000 cycles at 3 A g^(−1).展开更多
文摘背景与目的:弥漫大B细胞淋巴瘤(diffuse large B-cell lymphoma,DLBCL)的生发中心B细胞样(germinal center B-cell-like,GCB)亚型和非GCB(non-GCB)亚型在患者预后和治疗上存在差异,但目前依赖有创病理学检查。本研究基于多参数MRI构建影像组学和深度学习模型,旨在于术前无创性区分这两种亚型。方法:本研究回顾性分析2013年3月—2024年12月在复旦大学附属华山医院及外院经病理学检查确诊的DLBCL患者。使用多参数MRI扫描数据,结合4种影像组学机器学习[支持向量机(support vector machine,SVM)、逻辑回归(logistic regression,LR)、高斯过程(Gaussian process,GP)和朴素贝叶斯(Naive Bayes,NB)]和3种深度学习[密集连接卷积网络121(densely-connected convolutional networks 121,DenseNet121)、残差网络101(residual network 101,ResNet101)和高效网络B5(Efficient Net-b5)]建立DLBCL亚型分类模型。此外,两名经验不同的放射科医师在盲法下基于MRI图像独立分类DLBCL。模型和医师的诊断性能均通过接收者操作特征曲线下面积(area under the curve,AUC)、准确度(accuracy,ACC)和F1分数(F1-score,F1)等指标进行量化评估,以衡量其区分GCB和non-GCB亚型的能力。本研究经复旦大学附属华山医院伦理委员会批准(KY2024-663),所有患者均知情同意。结果:本研究共纳入173例患者(GCB型55例,non-GCB型118例)。影像组学和深度学习方法能有效地区分DLBCL亚型。其中,GP影像组学模型(基于T1-CE+T2-FLAIR+ADC序列)和DenseNet121深度学习模型(基于T1-CE+T2-FLAIR+ADC序列)表现最佳,在内部验证集上分别取得优异性能(GP:AUC=0.900,ACC=0.896,F1=0.840;DenseNet121:AUC=0.846,ACC=0.854,F1=0.774),并在外部验证集上保持稳健。并且,最优AI模型的分类效能优于经验丰富的放射科医师(医师最高AUC=0.678)。结论:基于多参数MRI特征的影像组学与深度学习模型可有效地鉴别DLBCL的GCB与non-GCB亚型。其中,GP与DenseNet121模型在处理复杂图像数据、特别是融合多序列特征组进行亚型分类时,呈现出优异的性能。
基金Project(52278380)supported by the National Natural Science Foundation of ChinaProject(2023JJ30670)supported by the National Science Foundation of and Technology Major Project of Hunan Province,China。
文摘This study proposes an alternative calculation mode for stresses on the slip surface(SS).The calculation of the normal stress(NS)on the SS involves examining its composition and expanding its unknown using the Taylor series.This expansion enables the reasonable construction of a function describing the NS on the SS.Additionally,by directly incorporating the nonlinear Generalized Hoke-Brown(GHB)strength criterion and utilizing the slope factor of safety(FOS)definition,a function of the shear stress on the SS is derived.This function considers the mutual feedback mechanism between the NS and strength parameters of the SS.The stress constraints conditions are then introduced at both ends of the SS based on the spatial stress relation of one point.Determining the slope FOS and stress solution for the SS involves considering the mechanical equilibrium conditions and the stress constraint conditions satisfied by the sliding body.The proposed approach successfully simulates the tension-shear stress zone near the slope top and provides an intuitive description of the concentration effect of compression-shear stress of the SS near the slope toe.Furthermore,compared to other methods,the present method demonstrates superior processing capabilities for the embedded nonlinear GHB strength criterion.
基金financial support from projects funded by the National Natural Science Foundation of China(52172038,22179017)the National Key Research and Development Program of China(2022YFB4101600,2022YFB4101601)。
文摘For rechargeable aqueous zinc-ion batteries(ZIBs),the design of nanocomposites comprised of electrochemically active materials and carbon materials with novel structures has great prom-ise in addressing the issue of electrical conductivity and structural stability in the electrode materials during electrochemical cycling.We report the production of a novel flexible electrode material,by anchoring MnO_(2) nanosheets on a B,N co-doped carbon nanotube ar-ray(BNCNTs)grown on carbon cloth(BNCNTs@MnO_(2)),which was fabricated by in-situ pyrolysis and hydrothermal growth.The generated BNCNTs were strongly bonded to the surface of the car-bon fibers in the carbon cloth which provides both excellent elec-tron transport and ion diffusion,and improves the stability and dur-ability of the cathode.Importantly,the BNCNTs offer more active sites for the hydrothermal growth of MnO_(2),ensuring a uniform dis-tribution.Electrochemical tests show that BNCNTs@MnO_(2) delivers a high specific capacity of 310.7 mAh g^(−1) at 0.1 A g^(−1),along with excellent rate capability and outstanding cycling stability,with a 79.7% capacity retention after 8000 cycles at 3 A g^(−1).
