中文摘要：

呼吸信号组成成分复杂,且是非平稳信号。对呼吸信号进行时域特征分析时,由于呼吸间隔提取存在误差,有时得到的特征值不能精确判断哮喘。为了提高诊断哮喘的精度,提出了检测哮喘的2个新的参数。首先对呼吸信号进行EMD（经验模态分解）分解,提出了一个参数——内禀模态能量比。然后对呼吸信号进行时域分析,利用HHT（Hilbert-Huang）变换和阈值法相结合的方法检测呼吸间隔,根据呼吸间隔符合指数定律,得到另一个参数。最后通过2个参数进行对比判别。通过典型数据分析结果表明,基于EMD分解的内禀模态能量比检测不受波形和基线漂移的影响,能够正确地反映呼吸信息,准确率高,可作哮喘病情评估随访的重要手段。

英文摘要：

Respiration signals consist of complicated composition and are non-stationary signals.When we do some feature analysis to respiration signals in the time domain,due to the error in interbreath interval extraction,the asthma can not be diagnosed accurately using the available characteristic values.To improve the diagnostic accuracy for asthma,this paper puts forward two new parameters for testing the asthma.First,EMD is employed to decompose the respiration signals and a new parameter,intrinsic mode energy ratio is given.Then,time domain analysis is performed on the respiration signals and the interbreath interval is detected using HHT combined with threshold method.Because interbreath interval follows the power law,we can get the other parameter.At last,comparing and distinguishing are carried out using these two parameters.The test result from typical data indicates that the method based on intrinsic mode energy ratio is not influenced by waveform and baseline shift;the proposed method can response respiration information more accurately and is an important means for the follow-up of asthma condition assessment.