This study presents a comprehensive analysis of 132 tornadic events in northeastern China from 2004 to 2023,utilizing radar and ERA5 reanalysis data to investigate the climatology,environmental drivers,and synoptic li...This study presents a comprehensive analysis of 132 tornadic events in northeastern China from 2004 to 2023,utilizing radar and ERA5 reanalysis data to investigate the climatology,environmental drivers,and synoptic linkages with Northeast China cold vortices(NCCVs)of tornadic storms under different convective modes.Results reveal that discrete storms account for 70%of events,with clustered cells(CC)being the most frequent mode,while significant tornadoes(EF2+)are primarily associated with isolated cells(IC)and broken lines(BL).The storm mode distribution in northeastern China resembles that of the central United States but with a higher proportion of CC and lower IC.In contrast,southern China exhibits a higher frequency of quasi-linear(QL)modes(>50%),similar to European patterns.Although no single parameter clearly differentiates between all tornado modes,distinct morphological characteristics emerge through specific parameter combinations:NL modes are characterized by high 0-1 km storm-relative helicity(SRH1)and humidity but low 0-6 km shear(SR6),whereas IC modes display contrasting features with low SRH1 and high CAPE.Notably,83%of tornadoes are associated with NCCVs,preferentially forming in southeastern/southwestern quadrants.Strong tornadoes favor southeastern quadrants,while NCCV intensity correlates with tornadic distance from vortex centers.Three characteristic synoptic configurations emerge:(T1)strong deep vortices with vertically aligned cold troughs,generating southeast-dominant tornado clusters characterized by a high proportion of BL and QL modes;(T2)weaker vortices featuring sub-synoptic troughs,with southern-distributed events dominated by a predominance of the CC mode;(T3)transverse-trough systems exhibiting CAPE-SRH decoupling and reduced tornadic activity.This study enhances our understanding of tornadoes in northeastern China,informing future research on formation mechanisms,prediction methods,and disaster prevention strategies.展开更多
BACKGROUND There is a lack of clinical evidence on the efficacy and safety of transitioning from a thrice-daily pre-mixed insulin or basal-prandial regimen to insulin degludec/aspart(IDegAsp)therapy,with insufficient ...BACKGROUND There is a lack of clinical evidence on the efficacy and safety of transitioning from a thrice-daily pre-mixed insulin or basal-prandial regimen to insulin degludec/aspart(IDegAsp)therapy,with insufficient data from the Chinese population.AIM To demonstrate the efficacy,safety,and treatment satisfaction associated with the transition to IDegAsp in type 2 diabetes mellitus(T2DM).METHODS In this 12-week open-label,non-randomized,single-center,pilot study,patients with T2DM receiving thrice-daily insulin or intensive insulin treatment were transitioned to twice-daily injections of insulin IDegAsp.Insulin doses,hemoglobin A1c(HbA1c)levels,fasting blood glucose(FBG),hypoglycemic events,a Diabetes Treatment Satisfaction Questionnaire,and other parameters were assessed at baseline and 12-weeks.RESULTS This study included 21 participants.A marked enhancement was observed in the FBG level(P=0.02),daily total insulin dose(P=0.03),and overall diabetes treatment satisfaction(P<0.01)in the participants who switched to IDegAsp.There was a decrease in HbA1c levels(7.6±1.1 vs 7.4±0.9,P=0.31)and the frequency of hypoglycemic events of those who switched to IDegAsp decreased,however,there was no statistically significant difference.CONCLUSION The present findings suggest that treatment with IDegAsp enhances clinical outcomes,particularly FBG levels,daily cumulative insulin dose,and overall satisfaction with diabetes treatment.展开更多
目的通过PET/CT评估弥漫大B细胞淋巴瘤(diffuse large B-cell lymphoma,DLBCL)患者化疗中期较化疗前动脉摄取18氟代脱氧葡萄糖(18F-fluorodeoxyglucose,18F-FDG)水平的变化,反映患者化疗过程中动脉炎症的变化情况。方法回顾性连续纳入2...目的通过PET/CT评估弥漫大B细胞淋巴瘤(diffuse large B-cell lymphoma,DLBCL)患者化疗中期较化疗前动脉摄取18氟代脱氧葡萄糖(18F-fluorodeoxyglucose,18F-FDG)水平的变化,反映患者化疗过程中动脉炎症的变化情况。方法回顾性连续纳入2021年1月—2023年1月于山东第二医科大学第一附属医院诊治的DLBCL患者作为研究对象。所有患者均在化疗前及化疗中期进行了18F-FDG PET/CT检查。使用最大靶本底比(maximum target-to-background ratio,TBR_(max))及平均靶本底比(mean target-to-background ratio,TBR_(mean))评估颈动脉和主动脉化疗前、化疗中期的动脉炎症水平。颈动脉评估部位包括颈动脉整体测量段(whole carotid artery,CA-WH)和颈动脉最严重病变段(most-diseased segment of the carotid,CA-MDS);主动脉评估部位包括升主动脉、主动脉弓、胸主动脉和腹主动脉。分别比较整体患者化疗中期和化疗前左、右侧CA-WH-TBR_(max)、CA-WH-TBR_(mean)、CA-MDS-TBR_(max)和CA-MDS-TBR_(mean),以及主动脉各部位的TBR_(max)和TBR_(mean)的差异。根据超声检查的颈动脉粥样硬化斑块情况,将患者分为不稳定斑块组、稳定斑块组及无斑块组,分别比较3组患者左、右两侧CA-WH-TBR_(max)和CA-WH-TBR_(mean)的Δ靶本底比(target-to-background ratio,TBR)(化疗中期TBR-化疗前TBR),评估3组化疗期间颈动脉炎症水平变化程度的差异。结果共纳入85例DLBCL患者,其中不稳定斑块组26例(30.6%),稳定斑块组24例(28.2%),无斑块组35例(41.2%)。整体患者中,化疗中期双侧CA-WH-TBR_(max)(左侧P=0.015,右侧P=0.034)、CA-MDS-TBR_(max)(左侧P<0.001,右侧P=0.041)、CA-MDS-TBR_(mean)(左侧P=0.029,右侧P=0.008)、升主动脉TBR_(max)(P=0.010)较化疗前显著增高,差异有统计学意义。不稳定斑块组左侧CA-WH-ΔTBR_(max)较无斑块组(P=0.005)更高,差异具有统计学意义;不稳定斑块组左侧CA-WH-ΔTBR_(mean)较无斑块组(P=0.003)、稳定斑块组(P=0.025)更高,差异均有统计学意义。结论化疗可加速DLBCL患者双侧颈动脉及升主动脉炎症的进展,PET/CT技术能够识别化疗相关动脉炎症的演变,从而为临床早期干预动脉粥样硬化提供依据。展开更多
伪装目标与背景具有高度的相似性,极易受背景特征混淆,导致边界信息难以分辨且提取目标特征困难。目前主流的伪装目标检测(COD)算法主要针对性研究伪装目标本身及其边界行,忽略了图像背景与目标的相互关系,在复杂场景下的检测结果不理...伪装目标与背景具有高度的相似性,极易受背景特征混淆,导致边界信息难以分辨且提取目标特征困难。目前主流的伪装目标检测(COD)算法主要针对性研究伪装目标本身及其边界行,忽略了图像背景与目标的相互关系,在复杂场景下的检测结果不理想。为了探索背景和目标的潜在联系,提出一种通过挖掘边界和背景检测伪装目标的算法——I2DNet(Indirect to Direct Network)。该算法由5个部分组成:编码器,处理初始原始数据;边界指导的特征提取和挖掘框架,通过特征处理和特征挖掘提取更多精细的边界特征;背景引导的潜在特征学习框架,通过多尺度卷积探索更多的显著特征,同时基于注意力设计混合注意力模块(HAM),增强对背景特征的强化选择;信息补偿模块(ISM),弥补在特征处理过程中损失的细节信息;多任务协同分割解码器(MCD)则高效融合不同任务和模块提取的特征,并输出最终的预测结果。在广泛使用的3个数据集上的实验结果表明,所提算法优于其他15个先进模型,尤其在CAMO数据集上的平均绝对误差指标下降至0.042。展开更多
基金supported by the National Natural Science Foundation of China(Grant No.42305013)Joint Research Project for Meteorological Capacity Improvement(Grant Nos.23NLTSQ002 and 24NLTSQ001)+2 种基金China Meteorological Administration Tornado Key Laboratory(Grant No.TKL202307)the China Meteorological Administration Youth Innovation Team Fund(Grant No.CMA2024QN05)a research project of the Chinese Academy of Meteorological Science(Grant No.2023Z019)。
文摘This study presents a comprehensive analysis of 132 tornadic events in northeastern China from 2004 to 2023,utilizing radar and ERA5 reanalysis data to investigate the climatology,environmental drivers,and synoptic linkages with Northeast China cold vortices(NCCVs)of tornadic storms under different convective modes.Results reveal that discrete storms account for 70%of events,with clustered cells(CC)being the most frequent mode,while significant tornadoes(EF2+)are primarily associated with isolated cells(IC)and broken lines(BL).The storm mode distribution in northeastern China resembles that of the central United States but with a higher proportion of CC and lower IC.In contrast,southern China exhibits a higher frequency of quasi-linear(QL)modes(>50%),similar to European patterns.Although no single parameter clearly differentiates between all tornado modes,distinct morphological characteristics emerge through specific parameter combinations:NL modes are characterized by high 0-1 km storm-relative helicity(SRH1)and humidity but low 0-6 km shear(SR6),whereas IC modes display contrasting features with low SRH1 and high CAPE.Notably,83%of tornadoes are associated with NCCVs,preferentially forming in southeastern/southwestern quadrants.Strong tornadoes favor southeastern quadrants,while NCCV intensity correlates with tornadic distance from vortex centers.Three characteristic synoptic configurations emerge:(T1)strong deep vortices with vertically aligned cold troughs,generating southeast-dominant tornado clusters characterized by a high proportion of BL and QL modes;(T2)weaker vortices featuring sub-synoptic troughs,with southern-distributed events dominated by a predominance of the CC mode;(T3)transverse-trough systems exhibiting CAPE-SRH decoupling and reduced tornadic activity.This study enhances our understanding of tornadoes in northeastern China,informing future research on formation mechanisms,prediction methods,and disaster prevention strategies.
基金Supported by CAMS Innovation Fund for Medical Sciences,No.2023-I2M-C&T-B-043National High Level Hospital Clinical Research Funding,No.2022-PUMCH-B-015+1 种基金CAMS Innovation Fund for Medical Sciences,No.2021-1-12M-002Beijing Municipal Natural Science Foundation,No.M22014.
文摘BACKGROUND There is a lack of clinical evidence on the efficacy and safety of transitioning from a thrice-daily pre-mixed insulin or basal-prandial regimen to insulin degludec/aspart(IDegAsp)therapy,with insufficient data from the Chinese population.AIM To demonstrate the efficacy,safety,and treatment satisfaction associated with the transition to IDegAsp in type 2 diabetes mellitus(T2DM).METHODS In this 12-week open-label,non-randomized,single-center,pilot study,patients with T2DM receiving thrice-daily insulin or intensive insulin treatment were transitioned to twice-daily injections of insulin IDegAsp.Insulin doses,hemoglobin A1c(HbA1c)levels,fasting blood glucose(FBG),hypoglycemic events,a Diabetes Treatment Satisfaction Questionnaire,and other parameters were assessed at baseline and 12-weeks.RESULTS This study included 21 participants.A marked enhancement was observed in the FBG level(P=0.02),daily total insulin dose(P=0.03),and overall diabetes treatment satisfaction(P<0.01)in the participants who switched to IDegAsp.There was a decrease in HbA1c levels(7.6±1.1 vs 7.4±0.9,P=0.31)and the frequency of hypoglycemic events of those who switched to IDegAsp decreased,however,there was no statistically significant difference.CONCLUSION The present findings suggest that treatment with IDegAsp enhances clinical outcomes,particularly FBG levels,daily cumulative insulin dose,and overall satisfaction with diabetes treatment.
文摘伪装目标与背景具有高度的相似性,极易受背景特征混淆,导致边界信息难以分辨且提取目标特征困难。目前主流的伪装目标检测(COD)算法主要针对性研究伪装目标本身及其边界行,忽略了图像背景与目标的相互关系,在复杂场景下的检测结果不理想。为了探索背景和目标的潜在联系,提出一种通过挖掘边界和背景检测伪装目标的算法——I2DNet(Indirect to Direct Network)。该算法由5个部分组成:编码器,处理初始原始数据;边界指导的特征提取和挖掘框架,通过特征处理和特征挖掘提取更多精细的边界特征;背景引导的潜在特征学习框架,通过多尺度卷积探索更多的显著特征,同时基于注意力设计混合注意力模块(HAM),增强对背景特征的强化选择;信息补偿模块(ISM),弥补在特征处理过程中损失的细节信息;多任务协同分割解码器(MCD)则高效融合不同任务和模块提取的特征,并输出最终的预测结果。在广泛使用的3个数据集上的实验结果表明,所提算法优于其他15个先进模型,尤其在CAMO数据集上的平均绝对误差指标下降至0.042。
