OBJECTIVE Lychee seed,a famous traditional Chinese medicine,recently were reported to improve the learning and memory abilities in mice.However,it is still unclear whether lychee seed saponins(LSS)can improve the cogn...OBJECTIVE Lychee seed,a famous traditional Chinese medicine,recently were reported to improve the learning and memory abilities in mice.However,it is still unclear whether lychee seed saponins(LSS)can improve the cognitive function and associated mechanisms.METHODS In present studies,we established the Alzheimer disease(AD)model by injecting Aβ25-35 into the lateral ventricle of rats.Then the spatial learning and memory abilities of LSS-treated rats were evaluated with the Morris water maze,meanwhile the protein expressions of AKT,GSK3β and Tau in the hippocampal neuron were analyzed by immunohistochemistry and Western blotting.RESULTS The results showed LSS can improve the cognitive functions of AD rats through shortening the escape latency,increasing the number across the platform,platform quadrant dwell time and the percentage of the total distance run platform quadrant.The protein expression of AKT was significantly up-regulated and that of GSK3β and Tau were decreased remarkably in the hippocampal CA1 area.CONCLUSION Our study is the first to show that LSS significantly improve the cognitive function and prevent hippocampal neuronal injury of the rats with AD by activation of the PI3K/AKT/GSK3βsignaling pathway,suggesting LSS may be developed into the nutrient supplement for the treatment of AD.展开更多
针对当前情绪脑电信号(emotion electroencephalogram,EM-EEG)识别研究中时间域信息的时间尺度难以把握和空间域信息易被忽视致使辨识率停滞不前,以及采集EM-EEG时通道过多导致信息冗余和信息处理成本增加等问题,提出了基于CNN的时-空...针对当前情绪脑电信号(emotion electroencephalogram,EM-EEG)识别研究中时间域信息的时间尺度难以把握和空间域信息易被忽视致使辨识率停滞不前,以及采集EM-EEG时通道过多导致信息冗余和信息处理成本增加等问题,提出了基于CNN的时-空卷积优化融合网络进行EM-EEG识别研究。该融合网络由提取EM-EEG时域信息的长卷积(long convolution,L-Conv)CNN和提取EM-EEG空域信息的CNN并联组成,在CNN模型时-空优化中使用粒子群算法(particle swarm optimization,PSO)对时域CNN中的L-Conv尺度进行了优化,并使用短时功率谱(short time power spectrum,STPS)的相关分析方法进行空域CNN模型通道数目优化,深层且有效地提取了EEG中的时间域和空间域特征。结果表明,提出的时-空卷积优化融合CNN在SEED IV数据集上对平和、悲伤、恐惧、高兴4种情绪最终准确率可以达到90.13%,相比传统单一CNN的识别准确率提高了4.76%,并且通道数目由62路降低至33路,缩减了46.77%,证实了本方法的可行性。展开更多
基金supported by Science and Technology Planning Project of Sichuan Province(2008SZ0050,14JC0798)Educational Commission of Sichuan Province(10ZA035,15ZA0155)+1 种基金Science and Technology Program of Luzhou(2015-S-43,2016LZXNYD-T03)Key Development Program of Southwest Medical University(2010ZD-010)
文摘OBJECTIVE Lychee seed,a famous traditional Chinese medicine,recently were reported to improve the learning and memory abilities in mice.However,it is still unclear whether lychee seed saponins(LSS)can improve the cognitive function and associated mechanisms.METHODS In present studies,we established the Alzheimer disease(AD)model by injecting Aβ25-35 into the lateral ventricle of rats.Then the spatial learning and memory abilities of LSS-treated rats were evaluated with the Morris water maze,meanwhile the protein expressions of AKT,GSK3β and Tau in the hippocampal neuron were analyzed by immunohistochemistry and Western blotting.RESULTS The results showed LSS can improve the cognitive functions of AD rats through shortening the escape latency,increasing the number across the platform,platform quadrant dwell time and the percentage of the total distance run platform quadrant.The protein expression of AKT was significantly up-regulated and that of GSK3β and Tau were decreased remarkably in the hippocampal CA1 area.CONCLUSION Our study is the first to show that LSS significantly improve the cognitive function and prevent hippocampal neuronal injury of the rats with AD by activation of the PI3K/AKT/GSK3βsignaling pathway,suggesting LSS may be developed into the nutrient supplement for the treatment of AD.
文摘针对当前情绪脑电信号(emotion electroencephalogram,EM-EEG)识别研究中时间域信息的时间尺度难以把握和空间域信息易被忽视致使辨识率停滞不前,以及采集EM-EEG时通道过多导致信息冗余和信息处理成本增加等问题,提出了基于CNN的时-空卷积优化融合网络进行EM-EEG识别研究。该融合网络由提取EM-EEG时域信息的长卷积(long convolution,L-Conv)CNN和提取EM-EEG空域信息的CNN并联组成,在CNN模型时-空优化中使用粒子群算法(particle swarm optimization,PSO)对时域CNN中的L-Conv尺度进行了优化,并使用短时功率谱(short time power spectrum,STPS)的相关分析方法进行空域CNN模型通道数目优化,深层且有效地提取了EEG中的时间域和空间域特征。结果表明,提出的时-空卷积优化融合CNN在SEED IV数据集上对平和、悲伤、恐惧、高兴4种情绪最终准确率可以达到90.13%,相比传统单一CNN的识别准确率提高了4.76%,并且通道数目由62路降低至33路,缩减了46.77%,证实了本方法的可行性。
