[Objectives]To determine the optimal preparation technology of Clerodendrum bungei Steud.extract gel by orthogonal test and gel quality test method in General Rule 0114 of Chinese Pharmacopoeia(Volume IV,2020 Edition)...[Objectives]To determine the optimal preparation technology of Clerodendrum bungei Steud.extract gel by orthogonal test and gel quality test method in General Rule 0114 of Chinese Pharmacopoeia(Volume IV,2020 Edition),and to study its anorectal pharmacodynamics and drug release in vitro.[Methods]Carbomer 940,propylene glycol and absolute ethyl alcohol were selected as the main factors,and the preparation technology of C.bungei Steud.extract gel was optimized by orthogonal test.The mouse model of ulcerative hemorrhoids was established with glacial acetic acid(HAC)and compared with Ma Yinglong musk hemorrhoids ointment.The recovery of trauma was compared between the two groups.At the same time,porcine small intestine was used as semi-permeable membrane to make diffusion cell to simulate anal environment,and the drug release in vitro was studied.[Results]The C.bungei Steud.extract gel was smooth in appearance and good in stability.It could effectively treat anal ulcer in mice and release quickly in vitro.[Conclusions]The formula is reasonable,and the effect of animal experiment is remarkable,which can provide a new treatment plan for ulcerative hemorrhoids.展开更多
科学基金在推动我国科学发展中发挥着重要作用。然而,当前基金管理中存在项目成果与立项内容不匹配的现象,亟须建立细粒度的匹配度评估机制,以完善科研项目评价体系。实现项目成果细粒度匹配的首要前提是项目关键要素抽取。已有研究实...科学基金在推动我国科学发展中发挥着重要作用。然而,当前基金管理中存在项目成果与立项内容不匹配的现象,亟须建立细粒度的匹配度评估机制,以完善科研项目评价体系。实现项目成果细粒度匹配的首要前提是项目关键要素抽取。已有研究实现了项目文本的句子级语步分类,但难以捕捉细粒度信息;面向论文的问题方法抽取通常局限于单一研究问题的识别,难以适配包含多子问题的科研项目。基于此,本文聚焦科研项目关键要素抽取任务,首先界定研究背景、研究问题、研究方法、研究目标和研究意义五类项目关键要素;在此基础上,提出一种基于大模型的项目要素自动抽取方法,分别运用零样本学习、单样本学习和微调三种策略,探索大模型在项目要素抽取任务中的适配能力。研究结果表明,微调策略下最优模型ROUGE-L(recall-oriented understudy for gisting evaluation-L)指标达到0.849,验证了微调模型在项目要素抽取任务上的实用性和有效性。本文抽取的项目要素能够直接服务于项目成果细粒度匹配场景,为后续下游任务的开展提供方法支持以及为科研项目管理的智能化提供支撑。展开更多
为解决现有道路裂缝检测模型精度低、易漏检误检等问题,提出了一种基于混合局部通道注意力机制的YOLOv10n道路裂缝检测模型。针对道路裂缝细长且不规则的形状特点,首先在模型特征提取网络引入混合局部通道注意力机制(mixed local channe...为解决现有道路裂缝检测模型精度低、易漏检误检等问题,提出了一种基于混合局部通道注意力机制的YOLOv10n道路裂缝检测模型。针对道路裂缝细长且不规则的形状特点,首先在模型特征提取网络引入混合局部通道注意力机制(mixed local channel attention,MLCA),增强模型对道路裂缝特征的提取能力;其次,采用重参数化泛化特征金字塔网络(reparameterized generalized feature pyramid network,RepGFPN)优化原始颈部网络,充分融合多尺度下的裂缝特征信息;最后使用Focaler-IoU替换CIoU损失函数,调整模型训练不同裂缝样本的权重,加快收敛速度。在RDD2022_China数据集上的实验结果表明,改进后的模型相较于原始YOLOv10n模型检测准确率提升4.4%,平均精度均值(mean average precision,mAP)提高2.9%。与其他主流目标检测模型相比,改进后的模型在准确率、召回率和计算成本等方面均展现出最佳性能,验证了本文方法在道路裂缝检测任务中的有效性和优越性。展开更多
基金Supported by National Natural Science Foundation of China(31671954)。
文摘[Objectives]To determine the optimal preparation technology of Clerodendrum bungei Steud.extract gel by orthogonal test and gel quality test method in General Rule 0114 of Chinese Pharmacopoeia(Volume IV,2020 Edition),and to study its anorectal pharmacodynamics and drug release in vitro.[Methods]Carbomer 940,propylene glycol and absolute ethyl alcohol were selected as the main factors,and the preparation technology of C.bungei Steud.extract gel was optimized by orthogonal test.The mouse model of ulcerative hemorrhoids was established with glacial acetic acid(HAC)and compared with Ma Yinglong musk hemorrhoids ointment.The recovery of trauma was compared between the two groups.At the same time,porcine small intestine was used as semi-permeable membrane to make diffusion cell to simulate anal environment,and the drug release in vitro was studied.[Results]The C.bungei Steud.extract gel was smooth in appearance and good in stability.It could effectively treat anal ulcer in mice and release quickly in vitro.[Conclusions]The formula is reasonable,and the effect of animal experiment is remarkable,which can provide a new treatment plan for ulcerative hemorrhoids.
文摘科学基金在推动我国科学发展中发挥着重要作用。然而,当前基金管理中存在项目成果与立项内容不匹配的现象,亟须建立细粒度的匹配度评估机制,以完善科研项目评价体系。实现项目成果细粒度匹配的首要前提是项目关键要素抽取。已有研究实现了项目文本的句子级语步分类,但难以捕捉细粒度信息;面向论文的问题方法抽取通常局限于单一研究问题的识别,难以适配包含多子问题的科研项目。基于此,本文聚焦科研项目关键要素抽取任务,首先界定研究背景、研究问题、研究方法、研究目标和研究意义五类项目关键要素;在此基础上,提出一种基于大模型的项目要素自动抽取方法,分别运用零样本学习、单样本学习和微调三种策略,探索大模型在项目要素抽取任务中的适配能力。研究结果表明,微调策略下最优模型ROUGE-L(recall-oriented understudy for gisting evaluation-L)指标达到0.849,验证了微调模型在项目要素抽取任务上的实用性和有效性。本文抽取的项目要素能够直接服务于项目成果细粒度匹配场景,为后续下游任务的开展提供方法支持以及为科研项目管理的智能化提供支撑。
文摘为解决现有道路裂缝检测模型精度低、易漏检误检等问题,提出了一种基于混合局部通道注意力机制的YOLOv10n道路裂缝检测模型。针对道路裂缝细长且不规则的形状特点,首先在模型特征提取网络引入混合局部通道注意力机制(mixed local channel attention,MLCA),增强模型对道路裂缝特征的提取能力;其次,采用重参数化泛化特征金字塔网络(reparameterized generalized feature pyramid network,RepGFPN)优化原始颈部网络,充分融合多尺度下的裂缝特征信息;最后使用Focaler-IoU替换CIoU损失函数,调整模型训练不同裂缝样本的权重,加快收敛速度。在RDD2022_China数据集上的实验结果表明,改进后的模型相较于原始YOLOv10n模型检测准确率提升4.4%,平均精度均值(mean average precision,mAP)提高2.9%。与其他主流目标检测模型相比,改进后的模型在准确率、召回率和计算成本等方面均展现出最佳性能,验证了本文方法在道路裂缝检测任务中的有效性和优越性。
