目的:探讨对自主乳头溢液性疾病,行乳腺导管造影后,进行全数字化乳腺摄影(full-field digital mammography,FFDM)与数字化乳腺断层(digital breast tomosynthesis,DBT)联合摄影,分析此方法在诊断乳头溢液性疾病中的应用价值。方法:选取...目的:探讨对自主乳头溢液性疾病,行乳腺导管造影后,进行全数字化乳腺摄影(full-field digital mammography,FFDM)与数字化乳腺断层(digital breast tomosynthesis,DBT)联合摄影,分析此方法在诊断乳头溢液性疾病中的应用价值。方法:选取我院2020年1月至2024年3月间进行乳腺导管造影且每一例都行FFDM+DBT联合摄影的乳头自主溢液患者116例,由两位医师分别对同一入组患者的图像进行回顾性阅片,一名医师只对FFDM图像进行阅片(对照组),另一名医师对FFDM联合DBT组合图像进行阅片(研究组),均对检出的病灶征象进行描述并做出影像诊断,进行以下两个方面的比较研究:(1)分别统计分析两组对导管内占位性病变的检出率。(2)统计分析两组诊断为导管扩张症的准确性。结果:(1)116例中,对照组(FFDM)检出导管内占位性病变49例,检出率为42.2%,研究组(FFDM+DBT)检出导管内占位性病变64例,检出率为55.2%,差异有统计学意义(P<0.005)。(2)116例中,对照组(FFDM)诊断为导管扩张症40例,研究组(FFDM+DBT)诊断为导管扩张26症例,其中对照组诊断为导管扩张症的40例中有14例在研究组DBT图像上通过逐层观察发现为导管重叠所致的假阳性。结论:(1)FFDM联合DBT导管造影对乳腺导管内占位性病变的检出率明显高于单纯FFDM乳腺导管造影。(2)FFDM联合DBT导管造影可避免单纯FFDM导管造影因多支导管重叠所致的假阳性,故前者诊断导管扩张症准确性优于后者。FFDM联合DBT导管造影乳头溢液性疾病的诊断中具有较高的应用价值。展开更多
A novel composite material(TD)composed of TS-1 microcrystalline and dendritic mesoporous silica nanospheres(DMSNs)was successfully prepared.The TD composite material had open pore structure and large specific surface ...A novel composite material(TD)composed of TS-1 microcrystalline and dendritic mesoporous silica nanospheres(DMSNs)was successfully prepared.The TD composite material had open pore structure and large specific surface area,which was conducive to the mass transfer of reactants and products.The Ti element in TS-1 could be used as an electron assistant,and the spillover d-electrons were conducive to the improvement of the sulfidation and dispersion of MoS_(2),thereby forming more type II MoS_(2) active phases.The incorporation of Ti could bring more Brønsted(B)and Lewis(L)acid,which was conducive to the hydrogenation pathway(HYD)selectivity(41.2%)of dibenzothiophene(DBT)hydrodesulfurization(HDS)and isomerization(ISO)route selectivity(21.9%)of 4,6-dimethyldibenzothiophene(4,6-DMDBT)HDS,thus improve the HDS activity of DBT and 4,6-DMDBT.NiMo/TD-70(Aging temperature=70℃)had the best HDS activities of DBT(99.0%)and 4,6-DMDBT(93.7%)due to its large open pore structure,good acidity,suitable metal-support interaction(MSI)and perfect dispersion of the metallic active sites.展开更多
Micro-mesoporous ZK-1 molecular sieves with different Si/Al ratios were used as supports for binary Co-Mo hydrodesulfurization（HDS） catalysts.The CoMo/ZK-1 catalysts were prepared using an over-loading impregnation ...Micro-mesoporous ZK-1 molecular sieves with different Si/Al ratios were used as supports for binary Co-Mo hydrodesulfurization（HDS） catalysts.The CoMo/ZK-1 catalysts were prepared using an over-loading impregnation method,and characterized using N2 physisorption,X-ray diffraction,temperature-programmed NH3 desorption,temperature-programmed reduction（TPR）,ultraviolet-visible diffuse reflectance spectroscopy,and high-resolution transmission electron microscopy（HRTEM）.The results show that the CoMo/ZK-1 catalysts have high surface areas（～700 m^2/g）,large pore volumes,and hierarchical porous structures,which promote the dispersion of Co and Mo oxide phases on the ZK-1 supports.The TPR results show that the interactions between the Co and Mo oxide phases and the ZK-1 support are weaker than those in the CoMo/γ-Al2O3 catalyst.The HRTEM results show that the CoMo/ZK-1 catalysts have better MoS2 dispersion and more active edge sites.The catalysts were tested in HDS of dibenzothiophene.Under mild reaction conditions,the activity of Co and Mo sulfides supported on ZK-1 was higher than those of Co and Mo sulfides supported on ZSM-5,A1KIT-1,and γ-Al2O3.展开更多
目的:探讨乳腺数字断层扫描(digital breast tomosynthesis,DBT)影像计算机辅助诊断方法在乳腺肿块良恶性鉴别中的价值。方法:本研究纳入我院乳腺癌早期筛查患者(2018年02月至2020年02月)190例,分别提取和筛选DBT影像中常规手工特征和...目的:探讨乳腺数字断层扫描(digital breast tomosynthesis,DBT)影像计算机辅助诊断方法在乳腺肿块良恶性鉴别中的价值。方法:本研究纳入我院乳腺癌早期筛查患者(2018年02月至2020年02月)190例,分别提取和筛选DBT影像中常规手工特征和深度学习神经网络特征,进行特征级融合,建立逻辑回归分类器模型,并绘制诺模图,通过ROC曲线、校正曲线和决策曲线对模型进行评估。结果:影像组学手工特征与深度学习特征融合筛选后得到含有三个特征的最优特征集,融合模型的诊断效果优于单独特征;无监督聚类分析和箱线图表明获得的特征具有良好的良恶性鉴别能力;基于融合特征构建的诺模图模型在训练集和测试集上获得AUC值分别为0.985(95%CI 0.956~1.000,敏感性=0.970,特异性=0.929)和0.984(95%CI 0.956~1.000,敏感性=0.909,特异性=0.931),绘制决策曲线(DCA)表明模型具有良好的潜在临床应用价值。结论:通过影像组学手工特征与深度学习特征融合并建立计算机模型可辅助临床医生提高乳腺DBT影像对肿块的良恶性鉴别能力。展开更多
文摘目的:探讨对自主乳头溢液性疾病,行乳腺导管造影后,进行全数字化乳腺摄影(full-field digital mammography,FFDM)与数字化乳腺断层(digital breast tomosynthesis,DBT)联合摄影,分析此方法在诊断乳头溢液性疾病中的应用价值。方法:选取我院2020年1月至2024年3月间进行乳腺导管造影且每一例都行FFDM+DBT联合摄影的乳头自主溢液患者116例,由两位医师分别对同一入组患者的图像进行回顾性阅片,一名医师只对FFDM图像进行阅片(对照组),另一名医师对FFDM联合DBT组合图像进行阅片(研究组),均对检出的病灶征象进行描述并做出影像诊断,进行以下两个方面的比较研究:(1)分别统计分析两组对导管内占位性病变的检出率。(2)统计分析两组诊断为导管扩张症的准确性。结果:(1)116例中,对照组(FFDM)检出导管内占位性病变49例,检出率为42.2%,研究组(FFDM+DBT)检出导管内占位性病变64例,检出率为55.2%,差异有统计学意义(P<0.005)。(2)116例中,对照组(FFDM)诊断为导管扩张症40例,研究组(FFDM+DBT)诊断为导管扩张26症例,其中对照组诊断为导管扩张症的40例中有14例在研究组DBT图像上通过逐层观察发现为导管重叠所致的假阳性。结论:(1)FFDM联合DBT导管造影对乳腺导管内占位性病变的检出率明显高于单纯FFDM乳腺导管造影。(2)FFDM联合DBT导管造影可避免单纯FFDM导管造影因多支导管重叠所致的假阳性,故前者诊断导管扩张症准确性优于后者。FFDM联合DBT导管造影乳头溢液性疾病的诊断中具有较高的应用价值。
基金This research was supported by the Science Foundation of China University of Petroleum-Beijing(No.2462023QNXZ002)the National Key R&D Program of China(2021YFA1501201)+2 种基金the National Natural Science Foundation of China(No.22278174)Independent research project of State Key Laboratory of heavy oil(2021–01)Shandong Excellent Young Scientists Fund Program(Overseas,2022HWYQ-082).
文摘A novel composite material(TD)composed of TS-1 microcrystalline and dendritic mesoporous silica nanospheres(DMSNs)was successfully prepared.The TD composite material had open pore structure and large specific surface area,which was conducive to the mass transfer of reactants and products.The Ti element in TS-1 could be used as an electron assistant,and the spillover d-electrons were conducive to the improvement of the sulfidation and dispersion of MoS_(2),thereby forming more type II MoS_(2) active phases.The incorporation of Ti could bring more Brønsted(B)and Lewis(L)acid,which was conducive to the hydrogenation pathway(HYD)selectivity(41.2%)of dibenzothiophene(DBT)hydrodesulfurization(HDS)and isomerization(ISO)route selectivity(21.9%)of 4,6-dimethyldibenzothiophene(4,6-DMDBT)HDS,thus improve the HDS activity of DBT and 4,6-DMDBT.NiMo/TD-70(Aging temperature=70℃)had the best HDS activities of DBT(99.0%)and 4,6-DMDBT(93.7%)due to its large open pore structure,good acidity,suitable metal-support interaction(MSI)and perfect dispersion of the metallic active sites.
基金financially supported by the National Natural Science Foundation of China (NNSFC,21206017)~~
文摘Micro-mesoporous ZK-1 molecular sieves with different Si/Al ratios were used as supports for binary Co-Mo hydrodesulfurization（HDS） catalysts.The CoMo/ZK-1 catalysts were prepared using an over-loading impregnation method,and characterized using N2 physisorption,X-ray diffraction,temperature-programmed NH3 desorption,temperature-programmed reduction（TPR）,ultraviolet-visible diffuse reflectance spectroscopy,and high-resolution transmission electron microscopy（HRTEM）.The results show that the CoMo/ZK-1 catalysts have high surface areas（～700 m^2/g）,large pore volumes,and hierarchical porous structures,which promote the dispersion of Co and Mo oxide phases on the ZK-1 supports.The TPR results show that the interactions between the Co and Mo oxide phases and the ZK-1 support are weaker than those in the CoMo/γ-Al2O3 catalyst.The HRTEM results show that the CoMo/ZK-1 catalysts have better MoS2 dispersion and more active edge sites.The catalysts were tested in HDS of dibenzothiophene.Under mild reaction conditions,the activity of Co and Mo sulfides supported on ZK-1 was higher than those of Co and Mo sulfides supported on ZSM-5,A1KIT-1,and γ-Al2O3.
文摘目的:探讨乳腺数字断层扫描(digital breast tomosynthesis,DBT)影像计算机辅助诊断方法在乳腺肿块良恶性鉴别中的价值。方法:本研究纳入我院乳腺癌早期筛查患者(2018年02月至2020年02月)190例,分别提取和筛选DBT影像中常规手工特征和深度学习神经网络特征,进行特征级融合,建立逻辑回归分类器模型,并绘制诺模图,通过ROC曲线、校正曲线和决策曲线对模型进行评估。结果:影像组学手工特征与深度学习特征融合筛选后得到含有三个特征的最优特征集,融合模型的诊断效果优于单独特征;无监督聚类分析和箱线图表明获得的特征具有良好的良恶性鉴别能力;基于融合特征构建的诺模图模型在训练集和测试集上获得AUC值分别为0.985(95%CI 0.956~1.000,敏感性=0.970,特异性=0.929)和0.984(95%CI 0.956~1.000,敏感性=0.909,特异性=0.931),绘制决策曲线(DCA)表明模型具有良好的潜在临床应用价值。结论:通过影像组学手工特征与深度学习特征融合并建立计算机模型可辅助临床医生提高乳腺DBT影像对肿块的良恶性鉴别能力。
基金Beijing Natural Science Foundation(2052009)Funding Project for Academic Human Resources Development in Institutions of Higher Learning Under the Jurisdiction of Beijing Municipality(PHR200907129)
