OBJECTIVE:To investigate the mechanism of electroacupuncture of sympathetic nerve activity and blood pressure reduction in the hypothalamic paraventricular nucleus(PVN)of spontaneous hypertensive rats(SHRs).METHODS:A ...OBJECTIVE:To investigate the mechanism of electroacupuncture of sympathetic nerve activity and blood pressure reduction in the hypothalamic paraventricular nucleus(PVN)of spontaneous hypertensive rats(SHRs).METHODS:A total of 64 male SHRs were divided into four groups:model,sham-operated(Sham),electroacupuncture(EA),and N-methyl-D-aspartate receptor antagonist and electroacupuncture(NRA+EA).In addition,16 Wistar-Kyoto rats were used as controls.PVN stereotaxic surgery was performed in both the Sham and NRA+EA groups,while the EA and NRA+EA groups received 14 d of electroacupuncture.Blood pressure(BP)and heart rate(HR)were measured the day before the intervention and every other day.After 14 d of intervention,the rats in each group were tested for renal sympathetic nerve activity(RSNA).The associated factor levels were determined using Western blotting,reverse transcription-polymerase chain reaction(RTPCR),enzyme-linked immunosorbent assay(ELISA)and immunofluorescence assays.RESULTS:In comparison to the model group,the EA and NRA+EA groups had significantly lower BP,HR,and RSNA(P<0.01).The expression of N-methyl-Daspartate receptor(NMDAR),angiotensin II(Ang II),angiotensin II type 1(AT1),tumor necrosis factor-α,interleukin-1β,norepinephrine and arginine vasopressin was significantly lower in the EA and NRA+EA groups(P<0.01).Moreover,the antihypertensive effect of NRA+EA group outperformed to the EA group.CONCLUSIONS:Electroacupuncture effectively reduced the BP and sympathetic nerve excitability in SHRs.The mechanism was linked to the inhibition of NMDARmediated Ang II/AT1 and the inflammatory response in PVN.展开更多
针对布匹瑕疵自动化检测,基于传统的机器视觉方法依赖于人工设计特征,对具有复杂背景图案的花色布瑕疵特征提取难度非常大,因此提出一种基于改进Faster R-CNN(faster region with convolutional neural network)的花色布瑕疵检测算法。...针对布匹瑕疵自动化检测,基于传统的机器视觉方法依赖于人工设计特征,对具有复杂背景图案的花色布瑕疵特征提取难度非常大,因此提出一种基于改进Faster R-CNN(faster region with convolutional neural network)的花色布瑕疵检测算法。在Faster R-CNN的基础上使用Resnet-50作为主干网络,嵌入可变形卷积来提高瑕疵特征的学习能力。通过设计多尺度模型来提高小瑕疵的检测,引入级联网络来提高瑕疵检测精度和定位准确度,构造优化的损失函数来降低样本不平衡影响。通过试验验证了该算法的有效性。结果表明,瑕疵检测效果准确率达94.97%,并能精准定位瑕疵位置,可满足工厂的实际需求。展开更多
针对现有轮廓匹配算法在轮廓缺失、轮廓噪声点、非线性光照下匹配效果差的问题,提出了基于方向梯度直方图(histogram of oriented gradient,HOG)特征描述的轮廓匹配算法。将轮廓曲线点离散化,计算所有离散点梯度并统计其在不同方向上的...针对现有轮廓匹配算法在轮廓缺失、轮廓噪声点、非线性光照下匹配效果差的问题,提出了基于方向梯度直方图(histogram of oriented gradient,HOG)特征描述的轮廓匹配算法。将轮廓曲线点离散化,计算所有离散点梯度并统计其在不同方向上的密度分布,以欧氏距离作为相似性度量,采用金字塔分层搜索策略进行搜索。统计所有离散点梯度信息得到梯度方向直方图,融合了轮廓的局部信息和全局信息,使得该算法对部分缺失、遮挡、边缘噪声具有较高的鲁棒性。以印花鞋面表面轮廓为试验对象,使用该算法进行匹配,结果表明,在复杂背景下,该算法对加入人为干扰的鞋面匹配效果较好,且采用金字塔分层搜索策略,匹配速度能满足实际应用的需要。展开更多
基金Doctoral Research Fund Project of Qilu Hospital of Shandong University(Qingdao):Mechanistic Study on the Improvement of Vertigo Caused by Posterior Circulation Ischemia by Acupuncture through Regulating Cerebral Blood Flow in Rats with Posterior Circulation Ischemia Vertigo(No.QDKY2023BS19)National Natural Science Foundation of China:From Microrna 9 Regulate P2X7 Receptor of Microglia in Paraventricular Nucleus of Hypothalamus to Explore the Effect of Electroacupuncture on Sympathetic Nerve Excitability in Spontaneously Hypertensive Rats(No.82074553)。
文摘OBJECTIVE:To investigate the mechanism of electroacupuncture of sympathetic nerve activity and blood pressure reduction in the hypothalamic paraventricular nucleus(PVN)of spontaneous hypertensive rats(SHRs).METHODS:A total of 64 male SHRs were divided into four groups:model,sham-operated(Sham),electroacupuncture(EA),and N-methyl-D-aspartate receptor antagonist and electroacupuncture(NRA+EA).In addition,16 Wistar-Kyoto rats were used as controls.PVN stereotaxic surgery was performed in both the Sham and NRA+EA groups,while the EA and NRA+EA groups received 14 d of electroacupuncture.Blood pressure(BP)and heart rate(HR)were measured the day before the intervention and every other day.After 14 d of intervention,the rats in each group were tested for renal sympathetic nerve activity(RSNA).The associated factor levels were determined using Western blotting,reverse transcription-polymerase chain reaction(RTPCR),enzyme-linked immunosorbent assay(ELISA)and immunofluorescence assays.RESULTS:In comparison to the model group,the EA and NRA+EA groups had significantly lower BP,HR,and RSNA(P<0.01).The expression of N-methyl-Daspartate receptor(NMDAR),angiotensin II(Ang II),angiotensin II type 1(AT1),tumor necrosis factor-α,interleukin-1β,norepinephrine and arginine vasopressin was significantly lower in the EA and NRA+EA groups(P<0.01).Moreover,the antihypertensive effect of NRA+EA group outperformed to the EA group.CONCLUSIONS:Electroacupuncture effectively reduced the BP and sympathetic nerve excitability in SHRs.The mechanism was linked to the inhibition of NMDARmediated Ang II/AT1 and the inflammatory response in PVN.
文摘针对布匹瑕疵自动化检测,基于传统的机器视觉方法依赖于人工设计特征,对具有复杂背景图案的花色布瑕疵特征提取难度非常大,因此提出一种基于改进Faster R-CNN(faster region with convolutional neural network)的花色布瑕疵检测算法。在Faster R-CNN的基础上使用Resnet-50作为主干网络,嵌入可变形卷积来提高瑕疵特征的学习能力。通过设计多尺度模型来提高小瑕疵的检测,引入级联网络来提高瑕疵检测精度和定位准确度,构造优化的损失函数来降低样本不平衡影响。通过试验验证了该算法的有效性。结果表明,瑕疵检测效果准确率达94.97%,并能精准定位瑕疵位置,可满足工厂的实际需求。
文摘针对现有轮廓匹配算法在轮廓缺失、轮廓噪声点、非线性光照下匹配效果差的问题,提出了基于方向梯度直方图(histogram of oriented gradient,HOG)特征描述的轮廓匹配算法。将轮廓曲线点离散化,计算所有离散点梯度并统计其在不同方向上的密度分布,以欧氏距离作为相似性度量,采用金字塔分层搜索策略进行搜索。统计所有离散点梯度信息得到梯度方向直方图,融合了轮廓的局部信息和全局信息,使得该算法对部分缺失、遮挡、边缘噪声具有较高的鲁棒性。以印花鞋面表面轮廓为试验对象,使用该算法进行匹配,结果表明,在复杂背景下,该算法对加入人为干扰的鞋面匹配效果较好,且采用金字塔分层搜索策略,匹配速度能满足实际应用的需要。
