Based on the vector diffraction theory, the effect of complex phase filters on intensity distribution of a radially polarized multi Gaussian beam in the focal region of high NA lens is theoretically investigated. It i...Based on the vector diffraction theory, the effect of complex phase filters on intensity distribution of a radially polarized multi Gaussian beam in the focal region of high NA lens is theoretically investigated. It is observed that a properly designed multi belt complex phase filter can generate subwavelength novel focal patterns including splitting of focal spots and generation of multiple focal spot segments such as eight, six and four focal spots along the optical axis are obtained. We expect that such an investigation is useful for optical manipulation and material processing, multiple high refractive index particle trapping technologies.展开更多
Polyester-based biodegradable polyurethane （PU） with different hard segment ratios was selected to modify the impact toughness of poly（L-lactide） （PLLA）. The influence of blending composition and hard segment ra...Polyester-based biodegradable polyurethane （PU） with different hard segment ratios was selected to modify the impact toughness of poly（L-lactide） （PLLA）. The influence of blending composition and hard segment ratio of PU on the phase morphology, crystallization behavior and mechanical properties of PLLA/PU blends has been investigated systematically. The results showed that the PU particles were uniformly dispersed in PLLA matrix at a scale from sub- microns to several microns. The glass transition temperature of PU within these blends decreased compared to that of neat PU, but rose slightly with its content and hard segment ratio. The presence of PU retarded the crystallization ability of PLLA, whereas enhanced its elongation at break and impact resistance effectively. As the PU content reaches up to 30 wt%, the phenomenon of brittle-ductile transition occurred, resulting in a rougher fracture surface with the formation of fibril-like structure. Moreover, under the same concentrations, the elongation at break and impact strength of PLLA blends decreased slightly with the increase of hard segment ratio of PU.展开更多
目的探讨不同训练集大小及不同增强期影像对肝癌患者增强CT影像自动分割效果的影响。方法选取2017年6月-2024年9月齐齐哈尔医学院附属医院和哈尔滨医科大学第四医院收治的100例肝细胞癌患者的增强CT影像数据。将数据按8︰1︰1比例随机...目的探讨不同训练集大小及不同增强期影像对肝癌患者增强CT影像自动分割效果的影响。方法选取2017年6月-2024年9月齐齐哈尔医学院附属医院和哈尔滨医科大学第四医院收治的100例肝细胞癌患者的增强CT影像数据。将数据按8︰1︰1比例随机分为训练集(80例)、验证集(10例)和测试集(10例)。为分析训练集大小对模型性能的影响,将训练集进一步划分为A组(80例)、B组(60例)、C组(40例)、D组(20例)和E组(10例);按增强期分为动脉期、门静脉期和延迟期三组。以E组作为基础模型,使用nnU-Net模型对各组影像进行自动分割,并比较不同组间的Dice相似系数(DSC)、95%Hausdorff距离(HD95)和交并比(Intersection over Union,IoU)。结果与B组、C组、D组和E组相比,A组(训练集80例)的分割效果最佳,DSC为0.82,IoU为0.71,HD95为62.09,差异具有统计学意义(P<0.05)。在不同增强期组中,动脉期组的分割效果优于静脉期组和延迟期组,DSC为(0.82±0.09),IoU为(0.71±0.13),HD95为(62.09±66.35),差异具有统计学意义(P<0.05)。结论采取较大训练集以及动脉期增强CT影像训练nnU-Net模型,可显著提高肝癌CT影像自动分割效果。展开更多
脑血管分割是脑血管疾病筛选、病理发展追踪和精准手术治疗的重要基础.受限于患者隐私保护和脑血管复杂的手工标注,现有的脑血管公开数据集样本少、人工标注信息不足,从而限制发展高性能的脑血管分割方法.为解决上述问题,本文提出了一...脑血管分割是脑血管疾病筛选、病理发展追踪和精准手术治疗的重要基础.受限于患者隐私保护和脑血管复杂的手工标注,现有的脑血管公开数据集样本少、人工标注信息不足,从而限制发展高性能的脑血管分割方法.为解决上述问题,本文提出了一种粗–精分割的Double U-Net协同优化模型(DU-Net),旨在提高小样本相位对比磁共振血管成像(Phase contrast magnetic resonance angiography,PC-MRA)脑血管分割数据集下的性能表现.首先,DU-Net的两部分路径骨干由3D U-Net构成,其中粗分割路径采用低通道数量的轻量卷积核,从而减少协同训练负担,实现脑血管拓扑特征的初步学习.接着,将拓扑形态传递到精分割路径,从而为模型补偿泛化特征.此外,在精分割路径的跳跃连接部分引入了压缩激励(Squeeze and excitation,SE)模块,为目标区域赋予更高的学习权重.实验结果显示,DU-Net在小样本脑血管数据集上的查准率、查全率、杰卡德系数和Dice系数分别是78.29%、67.91%、57.81%和72.59%,综合表现最优.结果表明,DU-Net具备更强的特征学习与识别能力,能够在小样本数据集上实现精确脑血管分割.展开更多
In this paper we first determine three phases of cell images： background, cytoplasm and nucleolus according to the general physical characteristics of cell images, and then develop a variational model, based on these...In this paper we first determine three phases of cell images： background, cytoplasm and nucleolus according to the general physical characteristics of cell images, and then develop a variational model, based on these characteristics, to segment nucleolus and cytoplasm from their relatively complicated backgrounds. In the meantime, the preprocessing obtained information of cell images using the OTSU algorithm is used to initialize the level set function in the model, which can speed up the segmentation and present satisfactory results in cell image processing.展开更多
Two series of polyurethanes based on mixed polychloromethyl methyl siloxane and poly-tetramethylene oxide in different weight fractions were synthesized. The phase separation ofsamples was studied using DSC and dynami...Two series of polyurethanes based on mixed polychloromethyl methyl siloxane and poly-tetramethylene oxide in different weight fractions were synthesized. The phase separation ofsamples was studied using DSC and dynamic mechanical property analysis. The results showedthat the introduction of chloromethyl group into polysiloxane increased its polarity and henceimproved the miscibilities with polytetramethylene oxide and polyurethane hard segment.Particularly, in the case of N-methyldiethanolamine extended materials, the surface and tensileproperties of these samples can be adjusted by various ratios of two soft segments.展开更多
文摘Based on the vector diffraction theory, the effect of complex phase filters on intensity distribution of a radially polarized multi Gaussian beam in the focal region of high NA lens is theoretically investigated. It is observed that a properly designed multi belt complex phase filter can generate subwavelength novel focal patterns including splitting of focal spots and generation of multiple focal spot segments such as eight, six and four focal spots along the optical axis are obtained. We expect that such an investigation is useful for optical manipulation and material processing, multiple high refractive index particle trapping technologies.
基金financially supported by the National Natural Science Foundation of China(No.51403210)China Postdoctoral Science Foundation(No.2014M550801)President Fund of University of Chinese Academy of Sciences(No.Y35102CN00)
文摘Polyester-based biodegradable polyurethane （PU） with different hard segment ratios was selected to modify the impact toughness of poly（L-lactide） （PLLA）. The influence of blending composition and hard segment ratio of PU on the phase morphology, crystallization behavior and mechanical properties of PLLA/PU blends has been investigated systematically. The results showed that the PU particles were uniformly dispersed in PLLA matrix at a scale from sub- microns to several microns. The glass transition temperature of PU within these blends decreased compared to that of neat PU, but rose slightly with its content and hard segment ratio. The presence of PU retarded the crystallization ability of PLLA, whereas enhanced its elongation at break and impact resistance effectively. As the PU content reaches up to 30 wt%, the phenomenon of brittle-ductile transition occurred, resulting in a rougher fracture surface with the formation of fibril-like structure. Moreover, under the same concentrations, the elongation at break and impact strength of PLLA blends decreased slightly with the increase of hard segment ratio of PU.
文摘目的探讨不同训练集大小及不同增强期影像对肝癌患者增强CT影像自动分割效果的影响。方法选取2017年6月-2024年9月齐齐哈尔医学院附属医院和哈尔滨医科大学第四医院收治的100例肝细胞癌患者的增强CT影像数据。将数据按8︰1︰1比例随机分为训练集(80例)、验证集(10例)和测试集(10例)。为分析训练集大小对模型性能的影响,将训练集进一步划分为A组(80例)、B组(60例)、C组(40例)、D组(20例)和E组(10例);按增强期分为动脉期、门静脉期和延迟期三组。以E组作为基础模型,使用nnU-Net模型对各组影像进行自动分割,并比较不同组间的Dice相似系数(DSC)、95%Hausdorff距离(HD95)和交并比(Intersection over Union,IoU)。结果与B组、C组、D组和E组相比,A组(训练集80例)的分割效果最佳,DSC为0.82,IoU为0.71,HD95为62.09,差异具有统计学意义(P<0.05)。在不同增强期组中,动脉期组的分割效果优于静脉期组和延迟期组,DSC为(0.82±0.09),IoU为(0.71±0.13),HD95为(62.09±66.35),差异具有统计学意义(P<0.05)。结论采取较大训练集以及动脉期增强CT影像训练nnU-Net模型,可显著提高肝癌CT影像自动分割效果。
文摘脑血管分割是脑血管疾病筛选、病理发展追踪和精准手术治疗的重要基础.受限于患者隐私保护和脑血管复杂的手工标注,现有的脑血管公开数据集样本少、人工标注信息不足,从而限制发展高性能的脑血管分割方法.为解决上述问题,本文提出了一种粗–精分割的Double U-Net协同优化模型(DU-Net),旨在提高小样本相位对比磁共振血管成像(Phase contrast magnetic resonance angiography,PC-MRA)脑血管分割数据集下的性能表现.首先,DU-Net的两部分路径骨干由3D U-Net构成,其中粗分割路径采用低通道数量的轻量卷积核,从而减少协同训练负担,实现脑血管拓扑特征的初步学习.接着,将拓扑形态传递到精分割路径,从而为模型补偿泛化特征.此外,在精分割路径的跳跃连接部分引入了压缩激励(Squeeze and excitation,SE)模块,为目标区域赋予更高的学习权重.实验结果显示,DU-Net在小样本脑血管数据集上的查准率、查全率、杰卡德系数和Dice系数分别是78.29%、67.91%、57.81%和72.59%,综合表现最优.结果表明,DU-Net具备更强的特征学习与识别能力,能够在小样本数据集上实现精确脑血管分割.
基金supported by the National Basic Research Program of China (Grant No. 2011CB707701)the National Natural Science Foundation of China (Grant No. 60873124)+2 种基金the Joint Research Foundation of Beijing Education Committee (GrantNo. JD100010607)the International Science and Technology Supporting Programme (Grant No. 2008BAH26B00)the Zhejiang Service Robot Key Laboratory (Grant No. 2008E10004)
文摘In this paper we first determine three phases of cell images： background, cytoplasm and nucleolus according to the general physical characteristics of cell images, and then develop a variational model, based on these characteristics, to segment nucleolus and cytoplasm from their relatively complicated backgrounds. In the meantime, the preprocessing obtained information of cell images using the OTSU algorithm is used to initialize the level set function in the model, which can speed up the segmentation and present satisfactory results in cell image processing.
文摘Two series of polyurethanes based on mixed polychloromethyl methyl siloxane and poly-tetramethylene oxide in different weight fractions were synthesized. The phase separation ofsamples was studied using DSC and dynamic mechanical property analysis. The results showedthat the introduction of chloromethyl group into polysiloxane increased its polarity and henceimproved the miscibilities with polytetramethylene oxide and polyurethane hard segment.Particularly, in the case of N-methyldiethanolamine extended materials, the surface and tensileproperties of these samples can be adjusted by various ratios of two soft segments.
