BACKGROUND Diaphragmatic paralysis is typically associated with phrenic nerve injury.Neonatal diaphragmatic paralysis diagnosis is easily missed because its manifestations are variable and usually nonspecific.CASE SUM...BACKGROUND Diaphragmatic paralysis is typically associated with phrenic nerve injury.Neonatal diaphragmatic paralysis diagnosis is easily missed because its manifestations are variable and usually nonspecific.CASE SUMMARY We report a 39-week-old newborn delivered via vaginal forceps who presented with tachypnea but without showing other birth-trauma-related manifestations.The infant was initially diagnosed with pneumonia.However,the newborn still exhibited tachypnea despite effective antibiotic treatment.Chest radiography revealed right diaphragmatic elevation.M-mode ultrasonography revealed decreased movement of the right diaphragm.The infant was subsequently diagnosed with diaphragmatic paralysis.After 4 weeks,tachypnea improved.Upon re-examination using M-mode ultrasonography,the difference in bilateral diaphragmatic muscle movement was smaller than before.CONCLUSION Appropriate use of M-mode ultrasound to quantify diaphragmatic excursions could facilitate timely diagnosis and provide objective evaluation.展开更多
针对传统空气质量预测模型收敛速度慢,精度低的问题,提出一种基于变分模态分解(variational mode decomposition,VMD)和蜣螂优化算法(dung beetle optimizer,DBO)优化长短期记忆网络(long short term memory,LSTM)的预测模型。首先,针对...针对传统空气质量预测模型收敛速度慢,精度低的问题,提出一种基于变分模态分解(variational mode decomposition,VMD)和蜣螂优化算法(dung beetle optimizer,DBO)优化长短期记忆网络(long short term memory,LSTM)的预测模型。首先,针对AQI原始数据具有大量噪声的问题,使用VMD方法对非平稳信号进行模态分解以降低噪声对预测结果的影响从而获得多个不同特征的模态分量;其次,针对LSTM靠人工经验调参存在一定局限性,利用DBO算法对LSTM模型参数进行优化;最后,对分解后的各个子序列使用LSTM模型预测,将各个子序列进行叠加得到最后的预测结果。实验结果表明,VMD对非平稳数据的分解有助于提高预测精度,VMD-DBO-LSTM模型的性能较其他模型均有不同程度的提高,该模型预测的均方根误差为4.73μg/m^(3),平均绝对误差为3.61μg/m^(3),拟合度达到了97.8%。展开更多
基金Supported by Sichuan Provincial Science&Technology Program,No.2022JDKP0040Sichuan Provincial Health Commission Program,No.21PJ168+1 种基金Deyang Municipal Science&Technology Program,No.2021SZZ068College-level Project of Chengdu University of Traditional Chinese Medicine,No.YYZX2021026 and No.YYZX2021020.
文摘BACKGROUND Diaphragmatic paralysis is typically associated with phrenic nerve injury.Neonatal diaphragmatic paralysis diagnosis is easily missed because its manifestations are variable and usually nonspecific.CASE SUMMARY We report a 39-week-old newborn delivered via vaginal forceps who presented with tachypnea but without showing other birth-trauma-related manifestations.The infant was initially diagnosed with pneumonia.However,the newborn still exhibited tachypnea despite effective antibiotic treatment.Chest radiography revealed right diaphragmatic elevation.M-mode ultrasonography revealed decreased movement of the right diaphragm.The infant was subsequently diagnosed with diaphragmatic paralysis.After 4 weeks,tachypnea improved.Upon re-examination using M-mode ultrasonography,the difference in bilateral diaphragmatic muscle movement was smaller than before.CONCLUSION Appropriate use of M-mode ultrasound to quantify diaphragmatic excursions could facilitate timely diagnosis and provide objective evaluation.
文摘针对传统空气质量预测模型收敛速度慢,精度低的问题,提出一种基于变分模态分解(variational mode decomposition,VMD)和蜣螂优化算法(dung beetle optimizer,DBO)优化长短期记忆网络(long short term memory,LSTM)的预测模型。首先,针对AQI原始数据具有大量噪声的问题,使用VMD方法对非平稳信号进行模态分解以降低噪声对预测结果的影响从而获得多个不同特征的模态分量;其次,针对LSTM靠人工经验调参存在一定局限性,利用DBO算法对LSTM模型参数进行优化;最后,对分解后的各个子序列使用LSTM模型预测,将各个子序列进行叠加得到最后的预测结果。实验结果表明,VMD对非平稳数据的分解有助于提高预测精度,VMD-DBO-LSTM模型的性能较其他模型均有不同程度的提高,该模型预测的均方根误差为4.73μg/m^(3),平均绝对误差为3.61μg/m^(3),拟合度达到了97.8%。
